L.;CL.OGY

$

(T/

MORPHOLOGY OF SPERMATOPHYTES

OTHER BOTANICAL WORKS
BY DR. JOHN M. COULTER.

PLANT RELATIONS.

A First Book of Botany.
i2mo. Cloth, $1.10.

PLANT STRUCTURES.

A Second Book of Botany.
I2mo. Cloth, $1.20.

PLANT STUDIES.

An Elementary Botany.

izmo. Cloth, $1.25.

PLANTS.

A Text-Book of Botany.
I2mo. Cloth, $1.80.

D. APPLETON AND COMPANY,

NEW YORK.

MORPHOLOGY
OF SPERMATOPHYTES

BY

JOHN M. COULTER, PH. D.

HEAD OF THE DEPARTMENT OF BOTANY IN THE UNIVERSITY OF CHICAGO

AND

CHARLES J. CHAMBERLAIN, PH. D.

INSTRUCTOR IN BOTANY IN THE UNIVERSITY OF CHICAGO

NEW YORK

D. APPLETON AND COMPANY
1901

6

COPYRIGHT, 1901,
BY D. APPLETON AND COMPANY.

PKEFACE

THIS book has grown out of a course of lectures, accompanied
by laboratory work, given for several successive years. The
course presupposes at least a year of work in general morphology,
and is intended to prepare the student for research in the mor-
phology of Seed-plants. As a consequence, the book brings to-
gether and organizes the very voluminous and scattered litera-
ture of the subject, points out and discusses the .problems, seeks
to unify a very confusing terminology, and at the same time
contributes no small amount of original observation and illus-
tration.

In developing the subject it has been thought best to present
first the facts connected with the different groups, organizing
these facts in as systematic and definite a way as possible, so
that the condition of knowledge in reference to any feature may
stand out clearly. Following the presentation of the separate
great groups, their comparative morphology, history, and phy-
logeny are discussed.

At the end of each chapter which discusses a great group, a
list of the works cited is given. This bibliography is not in-
tended to be exhaustive, but to include those works which con-
tain definite contributions of fact or of opinion. At the close
of the volume there will be found a fairly complete bibliography
of the important papers.

While in the very nature of things the great body of ma-
terial in a book of this kind must be derived from the work of
numerous observers, yet most of the ground has been traversed

vi MORPHOLOGY OF SPERMATOPHYTES

several times by the authors and their students, certain gaps
have been filled up, and an original point of view has been estab-
lished. So far as possible, the illustrations are original, and some
of the series are more complete than any heretofore published;
but illustrations have been reproduced whenever they seem neces-
sary to a full understanding of the subject. Those which have
been copied are fully credited, and may thus be distinguished
from those which are published for the first time in this book.
References to figures in the text have not been multiplied, upon
the assumption that those who use the book are in the habit of
consulting figures.

One of the chief perplexities in connection with the litera-
ture of the subject is the confusing and often misleading ter-
minology employed by various authors. Not only have different
names been applied to a single morphological structure, but fre-
quently the same name has been used for very different struc-
tures. This has resulted not only in a confusion of terms, but in
greatly obscuring a very simple and consistent morphology.
We have attempted to introduce uniformity into the terminology,
not by proposing new names, but by selecting from those already
current the ones which seem the most appropriate. It is to be
hoped that this will aid to a better understanding of morpho-
logical equivalents.

It may be well to call attention to the fact that we close
the history of the sporophyte with the appearance of the spore
mother cell, rather than with the appearance of the spore.
This has seemed to us to be the best defined line of demarcation
between the two generations, both on account of the reduction
division, and because preceding this division the mother cell
passes into a more or less prolonged resting condition. It cer-
tainly represents- the greatest break in the continuity of the life
history. In this sense, therefore, the sporophyte ends with the
appearance of the spore mother cell, and later the gametophyte
begins with the reduction division.

PREFACE

Vll

o attempt has been made to present the anatomy of Seed-
plants. This subject is too vast, and too far removed from our
general purpose to be included. The few anatomical details
which have been given are simply those which have to do with
the general structure of organs, and which are essential to any
morphological discussion.

The authors appreciate the fact that, in a certain sense, a
book of this nature is out of date as soon as it has left the press,
especially as it deals with groups which are being investigated at
the present time with remarkable activity. Doubtless, within a
few months papers will appear which would have been of great
service; but in so large a group this is likely to remain always
true, and the authors have felt that a concise summary of knowl-
edge at any stage of progress is useful in stimulating and guiding
further research.

JOHN M. COULTEK.

CHARLES J. CHAMBERLAIN.
THE UNIVERSITY OF CHICAGO,
January. 1901.

CONTENTS

PART I.— GYMNOSPERMS

CHAPTER PAGE

I. — CYCADALES 1

The vegetative organs 2

The stem 2

The leaf 7

The root 8

The spore-producing members 10

The microsporangium 10

The megasporangium 14

The gatnetophytes 19

The female gametophyte 19

The male gametophyte . . . ' 23

Fertilization 28

The embryo . 30

II. — GlNKGOALES 35

The vegetative organs 35

The stem 35

The leaf 36

The root 38

The spore-producing members 38

The microsporangium 38

The megasporangium 38

The gametophytes 41

The female gametophyte .41

The male gametophyte 42

Fertilization 45

The embryo 48

III. — CONIFERALES 51

The vegetative organs 51

The stem 51

The leaf 60

The root 63

The spore-producing members 64

The microsporangium 64

The megasporangium 67

The gametophytes 81

The female gametophyte 81

The male gametophyte 89

Fertilization . .95

The embryo . . 98

The classification of conifers 106

ix

x MORPHOLOGY OF SPERMATOPHYTES

CHAPTER PAGE

IV.— GNETALES 112

The vegetative organs 113

The stem 116

The leaf 118

The root 118

The spore-producing members 119

The microsporangium 119

The megasporangium 120

The gametophytes 124

The female gametophyte 124

The male gametophyte . 127

Fertilization 129

The embryo 130

V. — FOSSIL GYMNOSPERMS 134

Cordaitales 135

Bennettitales 142

Cycadales 150

Ginkgoales . . . ... . . . . . .150

Coniferales 151

VI. — COMPARATIVE MORPHOLOGY OF GYMNOSPERMS 153

The vegetative organs . . .153

The stem 153

The leaf 155

The root 155

The spore-producing members 156

The microsporangium 156

The megasporangium 157

The gametophytes 160

The female gametophyte 160

The male gametophyte . . . . ' . . . . 162

Fertilization 164

The embryo 165

VII. — THE PHYLOGENY OF GYMNOSPERMS 168

VIIL— GEOGRAPHIC DISTRIBUTION OF GYMNOSPERMS 175

Cycadales 175

Ginkgoales 176

Coniferales 176

Gnetales 178

SUMMARY OF LITERATURE CITED 180

APPENDIX 185

MORPHOLOGY OF SPERMATOPHYTES

PART I.-GYMNOSPERMS

CHAPTER I

^

CYCADALES

Ix the present flora nine genera of Cycads are recognized,
containing between seventy-five and one hundred species, and
distributed about equally between the oriental and occidental
tropics. The group is of special interest on account of its fern-
like characters, but knowledge of the most critical structures is
confined to comparatively few genera and species. With the
establishment of laboratories in regions where Cycads are in-
digenous, and with the increased attention paid to their culti-
vation, there is every reason to hope that the group will soon be
fairly well understood, so far as its living representatives are
concerned. The following account merely summarizes the
meager information at present available, which may or may not
find general application in the group.

The gross character which serves to distinguish Cycads from
other Gymnosperm groups is the combination of unbranched
stems with a terminal rosette of comparatively few large and
branched leaves, which gives to the columnar forms the habit of
tree ferns or palms (Figs. 1, 2, 3, 4). The most remarkable rec-
ondite character of the group, a character which it shares with
Ginkgo, is the occurrence of peculiar multiciliate male cells,
demonstrated as yet only for Cycas, Zamia, and Stangeria, but
probably common to the whole group.

1

^PERMATOPIIYTES

I. THE VEGETATIVE ORGANS

THE STEM

The columnar or tuberous stems, with no distinguishable
internodes, are usually completely invested by an armor of thick

jp^

>

leaf bases and scale leaves, some tuberous stems resembling
huge cones. Probably the greatest height is attained by species

CYCADALES

of Cycas, some of which carry the crown of leaves upon a col-
umnar trunk 3 to 4 meters, or sometimes even 12 meters
hia'h, while the Australian Cycas media reaches a height
of over 20 meters; but, in the majority of cases, the stem is
short and stocky, assuming the so-called tuberous form. In
fact, all the stems when young have the tuberous form, and the
difference between the two types consists in the one retaining
this form and the other becoming columnar. Branched forms
are rare in nature, but seem
to be more common in cultiva-
tion. AVe are informed that
wild plants are apt to branch
after they reach a certain age.
Until recently, the ac-
counts of the histology of the
stem have been derived in the
main from species of Cycas,
Enceplialartos, and Stangeria.
A historical resume of the sub-
ject was published by \Vors-
dell 21 in connection with his
work upon Macrozamia, the
substance of which is as fol-
lows: In 1829 Brongniart 1
gave some account of the anat-
omy of the stem of Cycas rer-
ohda, refuting the idea that
Cycadean stems are similar to
those of Monocotyledons, and
pointing out their true Gymno-
sperm character. He claimed,
however, that there is no phloem, and that xylem is the only
vascular tissue developed. In 1832 Yon Mohl 2 published an
account of the stem anatomy of Cycas and two species of En-
ceplialartos (referred at that time to Zamia, and known as Z.
latifolia and Z. liorrida). In one of the species of Encephalarios
(his Z. latifolia, but an African Enceplialarios not to be con-
fused with Z. latifolia Loddiges) he discovered the mesh work
of vascular bundles in the pith, and the fact that single bundles
traverse the pith rays and enter the cortex. In 1841 Miquel 3

Fro. 2. — Cycas media in the middle and at
the right; Cycas 3'ormanbyana at the
left.— After F. Vox MCLLER.

MORPHOLOGY OF SPERMATOPHYTES

published his great monograph of the group, and incidentally
touched upon the stem anatomy of several genera. In 1861
Mettenius 4 published an account of Cycadean stems which has
remained until recently the chief source of information about
them. He included Cycas revoluta, Encephalartos horrida,
Dioon edule, and Zamia muricata, finding in the last two genera
none of the secondary cortical bundles characteristic of the first
two. In 1885 Costantin and Morot 15 published an account of
the structure of the pericycle of Cycas Siamensis, claiming that
in it the cortical cambium zones have their origin. In 1890
Solms-Laubach 17 published an account of Stangeria paradoxa,

in which the vascular
bundles are traced from
the peduncles to their
juncture with the stem
system, and no evidence
of secondary cortical
bundles is found. In
1891 Strasburger 18
brought together the
knowledge of stem anat-
omy with reference to
Cycas, but did not in-
clude other genera. The
most important contribu-

FIG. Z. — titanqeria. paradoxa, staminate plant; tions to the knowledge of

stem anatomy since are
those of Worsdell 21- 30
on Macrozamia and Bowenia, and that of Scott 24 on the pedun-
cles of Cycads.

The structure common to all of the stems investigated is a
large pith surrounded by a thick cylinder of open collateral
bundles and broad pith rays, and a conspicuous cortical region.
The primary bundles are common, and the primary cambium is
either short-lived or somewhat persistent. In Zamia, Dioon,
and Stangeria no further structures occur to interfere with the
resemblance to an ordinary coniferous or dicotyledonous stem,
and secondary thickening, whatever it may amount to, is effected
.by the primary cambium.

A conspicuous feature of the structure of Cycas, Encephal-

about one fifteenth natural size. — From Warm-
ing-Potter's Handbook of Systematic Botany.

CYCADALES

artos, Macrozamia, and Bowenia, however, is the development
of a series of secondary cortical cauline bundles by successive sec-
ondary cambium cylinders. In these cases the primary cambium

is short-lived, and a series of secondary cambiums is organized in
the cortex, the first giving rise to a vascular cylinder as prominent
as the primary one, the second developing much smaller and

6 MORPHOLOGY OP SPERM ATOPHYTES

more widely separated bundles, and so with diminishing con-
structive power until the outermost cylinder is recognized only
by the appearance of a small bundle here and there. In Cycas
these cortical bundles are concentric, as are the bundles of the
peduncles and leaves of Cycas, Stangeria, Bowenia, and some
species of Zamia and Ceratozamia. Moreover, in Macrozamia
and Bowenia Worsdell discovered what may be called a tertiary
cambium, developed between the successive secondary cam-
biums, and giving rise to small intermediate bundles with re-
versed orientation — that is, with xylem directed toward the
xylem of the next outer bundles. This suggests that the origi-
nal structure of all the cortical bundles was concentric, and that
in the layers of concentric bundles the meristem of the inner
portion of each bundle became gradually less functional, until
in most cases the concentric bundles have become collateral.

Such facts as the cortical concentric bundles of Cycas, the
concentric bundles in the peduncles and leaves of several genera,
the evidence of incomplete concentric bundles in Macrozamia
and Bowenia, all point to an ancient stem type in which layers of
concentric bundles were developed. It is important to note, as
Scott 24 has suggested, that such a stem type is indicated in sev-
eral large extinct groups, as Lyginodendreae, Poroxyleae, Pro-
topityeae, and Medulloseae, representing the so-called Cycado-
filices, and that the living Cycads constitute only a special group
of what was once an extensive alliance. In Cycas, probably
the most primitive of the living genera, the concentric character
of the cortical bundles is retained ; in Macrozamia and Bowenia
remnants of the inner part of the concentric bundles exist ; while
in still other genera vestiges of the primitive organization once
common to the stem also survive only in peduncles. The fact
that the concentric bundle is a prominent feature of the Pterido-
phytes, and that it is the prevailing one in Paleozoic vascular
plants, emphasizes the general ancient as well as Pteridophyte
character of the Cycads.

In Encephalartos and Macrozamia a system of cauline bun-
dles is alao developed in the pith, forming a dense network trav-
ersing the pith in every direction, each bundle attended by a
mucilage canal in contact with the phloem. Certain smaller
bundles pass out through the pith rays, the xylem and phloem
strands joining the corresponding elements of the primary cyl-

(TCADALES

inder, and the mucilage canals passing on to join the cortical
canal system.

The xylem elements are of the general Gymnosperm charac-
ter. They are all tracheids, the first formed being spiral in
type and the later scalariform; while in the tracheids of the
secondary xylem the characteristic bordered pits appear, or
a mixture of scalariform and reticulate thickenings may be
developed, as in Zamia. Mucilage ducts are common in all
organs, and the pith is also remarkably rich in stored starch.

THE LEAF

The alternation of rosettes of scale leaves and foliage leaves
near the apex of the stem is a conspicuous feature. According
to Goebel, every year or every other year a rosette of foliage
leaves is expanded, in the center of which is a terminal bud

5.— Cross section of a leaf of

revoluta.

invested by numerous scale leaves under whose protection a new
rosette of foliage leaves slowly develops (Fig. 1). In the green-
house material we have examined there are no such regular peri-
ods of unfolding. The alternation is still more striking when
sporophylls are included, scale leaves, foliage leaves, and sporo-
phylls being the order of succession.

The scale leaves are brown and dry. The foliage leaves are
pinnate and leathery, and generally develop expanded, but in
2

8 MORPHOLOGY OF SPERMATOPHYTES

Cycas the pinnae are circinate, and in Zamia, C eratozamia, and
Stangeria the leaf axis is circinate. In general, two leaf traces
connect the vascular systems of the stem and leaf, sometimes
bending and passing almost horizontally through the cortex,
sometimes, as found by Worsdell 21 in Macrozamia, passing
tangentially close beneath the surface for some distance and
gradually turning inward to the stem system. Upon entering
the leaf the two bundles break up variously, and the vein system
is characterized by the comparatively small amount of anasto-
mosing and the frequency of dichotomy, as in Encephalartos.

The anatomical structure of the leaves is quite similar to
that of the broad leaves of certain Conifers (Fig. 5). The
epidermis is strongly cuticularized, and the stomata occur only
on the lower surface, the guard cells being deeply sunken. The
leathery texture is due to a strong hypodermal development of
thick-walled, almost fibrous cells, beneath which is the display
of mesophyll. A distinct palisade region is organized, but the
colorless middle region, in which the veins run, is peculiar. It
consists of cells elongated transversely to the leaf axis and parallel
with the surface, and large intercellular spaces, and may repre-
sent the " transfusion tissue " of Yon Mohl. The absence of chlo-
rophyll and the elongation of the cells has suggested a conducting
region to compensate for lack of the branching and anastomosing
of veins. The same structure occurs among Conifers in Sciado-
pitys and other broad-leaved forms ; while in the needle leaves
the regular transfusion tissue, more definitely differentiated, is
within the bundle sheath.

THE EOOT

The primary root continues as a tap root, but numerous
branches and secondary roots extend in every direction, the
latter often extending upward and reaching the surface. An
interesting phenomenon connected with these upward rising
root branches, which often spread in feltlike masses about the
base of the stem in species of Cycas, Macrozamia, Ceratozamia,
Dioon, and other genera, is the occurrence upon them, usually
at the tip, of short " coral-like " branches, usually in clusters
(Fig. 6). These peculiar branches have been called " tuber-
cles," but they do not have the structure of the bodies which
receive that name in certain Leguminosae, being much modified

CYCADALES

9

and tuberclelike rootlets. The apparently dichotomous branch-
ing of these bodies has led to the general statement that the sec-
ondary roots of Cycads
branch dichotomously,
but the dichotomy even
in this case is only ap-
parent, for the true
apex, although com-
pletely checked in ac-
tivity, is evident be-
tween the branches. In
1872 Keinke 6 an-
nounced the presence of
an endophytic alga in
these tuberclelike root-
lets, which he referred
to Anabaena, and in
1894 Schneider19 pub-
lished an account which
gave much fuller de-

FIG. 6. — The so-called " root tubercles " of Cycas

., -.^ revoluta; about natural size. — After LIFE.

tails. Irom a recent

paper by A. C. Life,31 the following facts are obtained.
It appears that the soil about the roots is full of low algal and

fungous forms, and these
forms are found also thickly
clothing the roots. At the
very inception of a branch
bacterioid forms effect an
entrance, and are found in
abundance in the cells of the
apical region. Presumably
through their activity ex-
crescent growth begins, and
a definite zone of cortical
cells is disorganized, result-
ing in a cortical chamber
about midway in the cortex
(Fig. 7). Upon the surface

numerous lenticel-like openings are developed, and connect with

the cortical chamber, which soon becomes filled with Nostoc

FIG. 7. — Cross section of a " root tubercle "
of Cycas revoluta. showing the algal zone.
— After LIFE.

10 MORPHOLOGY OF SPERMATOPHYTES

forms. That the form is Anabaena, and that there is but a sin-
gle species, are both doubtful statements. The cortical cells ad-
jacent to the chamber send out papillate prolongations into it,
giving in cross section the appearance of cells radially elongated.

That the Algae are not the inciting cause of the excrescent
growth is apparent from the fact that it begins before the Algae
enter, and also from the fact that in some of the tuberclelike
rootlets Algae do not occur. The relations existing between the
fungous and algal elements without and within the plant, and the
root system, with its response in excrescent growth, lenticels,
cortical chamber, and internal cortical papillae, must be very
complex.

The root cap is said by De Bary 13 to arise by the splitting
off of the outer layers of the periblem which covers the meri-
stematic region, there being no true calyptrogen or dermatogen.
This character is asserted by De Bary to be common to all Gym-
nosperms, in which they differ decidedly from Angiosperms.

II. THE SPORE-PRODUCING MEMBERS

THE MICKOSPORANGITJM

— Oycads are dioecious, and the staminate strobilus in Cycas,
at least, is terminal; but its true position in the other genera,
although apparently terminal, remains in doubt. Strobili may
occur singly or several together, and in the latter case some at
least are certainly lateral. The numerous sporophylls are
arranged in a close spiral (Figs. 8, 9).

The sporophylls are persistent, often becoming very hard,
and compared with those of other groups are remarkably large.
In certain genera, as Cycas, Macrozamia, Ceratozamia, Stan-
geria, etc., they have the form of ordinary heavy scales. The
horizontal portion is narrowed at the insertion and broadens
outward, bearing beneath the very numerous sporangia, some-
times as many as one thousand ; while it is continued into a
more or less vertically expanded and sterile terminal region,
which is usually hairy without. In other genera, as Zamia,
the horizontal and vertical portions of the sporophylls become
more distinctly differentiated, the former becoming a slender
stalk, the latter a peltate expansion beneath which the sporangia
are clustered, as in Equisetum. The sporangia are scattered,

CYCADALES

11

or they may occur in definite groups of two to five, which may
be called sori, as in Cycas, Stangeria, and Zamia (Fig. 9).

The first full account of the development of the microspo-
rangium was published by Treub J1 in 1881, who investigated
especially Zamia muricata. The latest investigation is that of
Siangeria paradoxa by Lang,23 who has confirmed Treub's re-
sults in a general way, and has
cleared up certain points which were
doubtful.

The first indication of a sporan-
gium is the differentiation of a
hypodermal plate of cells (four in
^tnngeria), which constitutes the
archesporium (Fig. 10). This arche-
sporium has been homologized with
the hypodermal plate of cells which
appears in the development of the
microsporangia of heterosporous
Pteridophytes. It should be remem-
bered, however, that the periblem
always gives rise to the archespori-
um, and that this region, in the ab-
sence of a dermatogen, is superfi-
cial in the Pteridophytes. It would
seem to follow, therefore, that a
true homology would regard the
archesporium as superficial in Pteri-
dophytes and hypodermal in Sper-
matophytes. In presenting Treub'a
results for Zamia, in which the
archesporium is spoken of as a hypo-
dermal group of cells, Goebel 16 re-
marks that " there is without doubt a unicellular archesporium."
In Stangeria, however, Lang was unable to trace the plate of four
cells to a single one, and a one-celled archesporium for the micro-
sporangia of Cycads still remains to be proved. Each cell of
the archesporial plate divides by a periclinal wall into a tabular
outer and a larger inner cell (Fig. 10). The outer daughter
plate is sterile, and there is no evidence that it contributes to
the sporogenous tissue, as in some Pteridophytes. It is com-

FIG. 8. — Zamia muricata : A, stam-
inate strobilus, two thirds nat-
ural size ; B, transverse section
of A ; C", single peltate stamen
showing attachment of spo-
rangia ; /), upper part of ovu-
late strobilus, two thirds natu-
ral size ; E, transverse section
of D ; F, longitudinal section
of a ripe seed. — After KAR-
STEX.

12

MORPHOLOGY OF SPERMATOPHYTES

monly spoken of as the " primary tapetum," but since it does
not always give rise to the tapetum it would seem better to refer
to it as primary wall cells of the sporangium. By continued

FIG. 9. — Microsporophylls of Cycads : A, Cycas circinalis ; a, entire sporophyll, showing
sporangia on under surface ; 6, four groups of sporangia ; c, same as ft, after the
pollen has been shed. B, Zamia in tegri folia ; a, entire sporophyll ; b and c, clusters
of sporangia. — Ab and Ac after BLUME, the other illustrations after KICHARD.

periclinal divisions a wall consisting of three to six layers
(Stangeria) of cells is organized beneath the epidermis.

The larger cells of the inner daughter plate of the first divi-
sion of the archesporial cells are the primary sporogenous cells
(Fig. 10), which by division organize a large sporogenous mass.
Lang states that in 8tangeria, after a considerable mass of
sporogenous cells is formed, the tapetum is organized by the
sporogenous tissue, either by cutting off peripheral tabular
cells or by direct conversion of the peripheral cells. If this
be true, the functional tapetum is derived from sporogenous
rather than wall tissue. The tapetal cells are distinguished in
the usual way by their size and deeply staining contents, and in
a single layer completely invest the sporogenous mass. The
tapetum persists as a distinct layer, but its walls disappear, and
its nuclei often fragment. In its development two of the adja-
cent wall layers are disorganized and appear as a thin band
about it, contributing of their material to the tapetum, and in
this sense functioning as a tapetum.

With the complete organization of the tapetum division in
the sporogenous tissue ceases, and the cells become mother cells.
Tapetal work, which seems to be merely the contribution of

CYC AD ALES

13

nutrition to functioning mother cells by those which are adja-
cent, whether they be morphologically wall cells, sporogenous
cells, or even other mother cells, in Cycads is participated in
also by some of the mother cells, especially those next to the
distinctly organized tapetal layer. The functional mother
cells become isolated, and with their complete organization the
history of the sporophyte ends, the mother cells entering upon
that more or less prolonged resting period preparatory to the
reduction division and the beginning of the gametophyte.

The sporangia dehisce by means of a slit which extends
from the base along the side aw.ay from the sporophyll and ends
in a peculiar group of cells at the apex, whose morphological sig-

FIG. 10. — Stangeria paradoxa, showing development of microsporangium : A, section
showing two of the four cells forming the archesporial plate ; J?, each cell of the
plate has divided into an inner primary sporogeuous cell (shaded) and an outer pri-
mary wall cell ; f, a more advanced stage ; Z>, a still more advanced stage, showing
the tapetum (shaded) and the sporogenous cells (*) ; E, a mature sporangium con-
taining microspores, showing the tapetum (£) and the crushed cells of the wall (w} :
A-D. x 266 ; E, x 66.— After LAXG.

nificance is in question. The arrangement of thin and thick
walled regions in the sporangium wall, and the mechanism of
dehiscence, are said to be essentially as found in Angiopteris.

14 MORPHOLOGY OF SPERMATOPIIYTES

It is of interest to note that numerous stomata are found on
the sporangium wall of Cemtozamia, Encephalartos, and Stan-
geria, and doubtless in other genera. They are found on the
wall which is opposite to the slit of dehiscence — that is, toward
the sporophyll. Stomata also occur upon pollen sacs of certain
Angiosperms, but so far as recorded not upon the sporangia of
Pteridophytes.

The development of the sporangia, their large output of
spores, their arrangement in sori, their several-layered wall, and
their dehiscence, are all characters suggestive of Marattiaceae;
while the chief resemblance to Angiosperms is found in the
hypodermal archesporium.

THE MEGASPORANGIUM

In most cases a definite ovulate strobilus is organized, which
is terminal in Cycas, and possibly so in the other genera (Figs.
11, 12, 13). In Cycas the axis of the strobilus continues

FIG 11.-

* revoluta, ovulate strobilus.— From photograph taken in Lincoln Park,
Chicago.

growth after the seeds have been formed, produces foliage
leaves, and finally organizes another strobilus. In all other
cases, including the staminate strobilus of Cycas, a strobilus
stops the growth of the axis. The size of the ovulate strobilus

CYCADALES

15

\

of Cycads is worthy of note. In the parks of Chicago we ob-
tained a strobilus of Dioon edule which measured 30 centime-
ters in height and 20 centimeters in diameter ; and one of Cycas
revoluia which measured 50 cen-
timeters in height, and about
the same in diameter.

In Cycas the sporophylls re-
semble the foliage leaves in gen-
eral form (Fig. 14), the strobilus
appearing as a rosette of much
smaller and yellowish leaves.
An interesting transition may
be traced from the leaflike spo-
rophyll of Cycas re valuta to
such specialized ones as those of
Zamia or Ceraiozamia (Figs.
14, 15, 16). In the sporophylls
of Cycas re valuta the upper
pinnae are prominent, but the
lowest ones, usually three on
each side, are replaced by spo-
rangia. In Cycas circinalis the
upper sterile pinnse are much re-
duced; while in Cycas Norman-
byana the terminal portion of
the sporophyll is merely toothed, the lowest pair of teeth being
replaced by sporangia. In Dioon and Encephalartos the sterile
portion is still leaflike, but it is entire, and a pair of sporangia
occupy a basal position; while in Zamia and Ceraiozamia the
sterile tip has become a peltate expansion, and beneath it is
the pair of sporangia.

Investigations in reference to the development of the mega-
sporangium have been incidental to a study of the embryo sac
structures. Important among these contributions have been
those of Warming 8' 9 on Cycas in 1877 and 1879; of Treub llj 14
on Ceraiozamia and Zamia in 1881 and 1884; and most recently
of Lang 29 on Stangeria paradoxa. The authors have also had
the privilege of looking over material of Zamia kindly supplied
by Mr. H. J. Webber. From these accounts, and from per-
sonal observation, the general facts seem to be as follows:

FIG. 12. — Dioon edule, ovulate strobil-
us; surface view; x %. — From pho-
tograph taken iu Lincoln Park, Chi-
cago.

16

MORPHOLOGY OP SPERMATOPHYTES

The first indication of a megasporangium is the differentia-
tion of a group of sporogenous cells immediately beneath the
epidermis. It is not a hypodermal plate, as in the microsporan-
gium, but a hypodermal mass of considerable extent. Whether
this group represents the archesporium, or whether it can be
traced back to a few-celled or even to a one-celled archesporium,
is still in doubt. Our impression is that there is a many-celled
archesporium. Following the differentiation of sporogenous
tissue, the exterior sterile cells begin to divide rapidly, organ-
izing the large sterile apical region of the nucellus. About this
there develops a very thick integument with a long and narrow
micropyle.

By means of this development of the exterior sterile tissue
the sporogenous tissue becomes deeply placed in the nucellus,

FIG. IS. — Dioon edule, sectional view of strobilus represented in Fig. 12.

appearing almost at its very base. In the sporogenous tissue
usually one centrally placed cell shows its selection for func-
tioning by enlarging at the expense of adjacent cells. From its
subsequent behavior it seems that this enlarging cell is the

CYCADALES

17

mother cell, which closes the history of the sporophyte in this
direction, and later through a reduction division begins the
gametophyte generation.

Lang observed in Stangeria that about this mother cell
and its product a persistent layer of the sporogenous tissue

FIG. 14. — Cycas rerohita, ovulate sporophyll ;
one half natural size. — From photograph
taken in Washington Park, Chicago.

FIG. 15. — Dioon edule, ovulate sporo-
phylls from strobilus shown in Fig.
12; one half natural size.

organizes, apparently functioning as a tapetum. Whether a
definite layer be organized or not, the adjacent cells continue
to contribute nutrition to the mother cell and its functioning
spore.

During these changes in the sporogenous tissue, the apex of
the nucellus develops a more or less prominent beaklike process
which projects into the micropyle (Fig. 17). The cell walls
in this beak region become firm, and the whole structure forms a
persistent cap, very noticeable in Cycads, in GinJcgo, and in
such fossil forms as the Cordaitales and Bennettitales. Within
this firm beak the pollen chamber is organized, which may be
narrow or broad, but in any event is able to contain numerous
pollen grains, one or two dozen having been observed in a single
pollen chamber of Zamia.

18

MORPHOLOGY OF SPERMATOPHYTES

Between the base of the beak and the wall of the embryo sac
the micellar tissue is loose, and at the time of the development

FIG. 16. — Ovulate sporophylls of various Cycads : A, Cycas revoluta ; B, Cycas circin-
alis\ C, Cycas Normanbyana; D,Dioonedule\ E, Encephalartos Preissii \ F. Zamia
integrifolia; 6r, Ceratozamia Mexicana. — A, after SACHS; C, after F. VON MCLLER;
E, after MIQUEL: F, after RICHARD; B, D, G-, drawn for ENGLER and PRANTL'S Nat.
Pflanzenfam., from whicli the entire plate is taken.

of the pollen tubes this whole region of the nucellus disorgan-
izes, leaving more or less of a cavity between the pollen chamber
and the embryo sac. Into this cavity numerous pollen tubes
are often seen dangling from the pollen chamber (Fig. 18), a

FIG. 17. — Stangeria paradoxa, longitudi-
nal section of ovule, x 7 : <?, embryo
sac containing endosperm ; «, layer of
" sporogenous tissue " around the em-
bryo sac; JP, pollen chamber. — After
LANG.

FIG. 18. — Zamia integrifolia \ longitudi-
nal section of upper end of nucellus,
showing pollery tubes growing down
into the archegonial chamber : pt, pol-
len tubes ; pc, pollen chamber ; pg, rem-
nant of pollen grain. — After WEBBER.

fact especially evident when one lifts off the caplike micellar
beak, which from the disorganization of the tissue beneath be-
comes quite loose.

CYCADALES

19

III. THE GAMETOPHYTES

THE FEMALE GAMETOPHYTE

The gametophyte generation begins with the reduction divi-
sion of the enlarged mother cell which lies deeply imbedded in
the nucellus and surrounded by more or less sporogenous tissue
which does not function as such. According to Treub n and
Lang,29 in Ceratozamia and Stangeria the transverse division
of the mother cell results in a row of three cells (Fig. 19), which
may be called potential megaspores, the lowest of which becomes
the functional megaspore. Whether this is true of all Cycads
or not remains to be seen. The
selected megaspore enlarges
rapidly at the expense of the
two functionless megaspores,
disorganizing them as well as
a varying amount of adjacent
tissue. The outer wall of the
megaspore is said to become cu-
tinized in Cycas, and this may
be true of other genera as well.
In any event, it suggests deriva-
tion from forms in which the
spores are discharged.

The germination of the meg-
aspore and the development of
the female gametophyte was
first published in any detail by

Warming, 8j 9 Cycas circinalis being the form studied. Addi-
tional incidental testimony has been given by Treub,11 by
Ikeno 2S in his study of Cycas revoluta, and by Lang 29 in his
study of Stangeria pamdoxa. It is often stated in a general
way that the development of the endosperm resembles that in
Selaginella and Isoetes, but since our knowledge of the develop-
ment in these forms has been obtained by putting together
fragments of information, such a statement does not rep-
resent any very exact knowledge. Unpublished results ob-
tained by Miss F. M. Lyon with Selaginella show that some-
thing more than inference from fragments is needed. So far

FIG. 19. — Development of the megaspore :
A, Stangeria paradoxa, showing the
functional megaspore enlarging at the
expense of two functionless mega-
spores, x 250; i?, Ceratozamia longi-
folia, showing the three potential
megaspores, the lowest beginning to
enlarge, x 266. — A, after LANG ; B,
after TREUB.

20 MORPHOLOGY OF SPERMATOPHYTES

as the information goes, the sequence of events seems to be
as follows:

The nucleus of the megaspore divides, and free nuclear
division continues until a number of nuclei are formed, all of
the divisions being simultaneous, as evidenced by the fact that
all the nuclei in the embryo sac show the same karyokinetic
phase. Early in these divisions the free nuclei pass to the wall
of the sac and become imbedded in a parietal cytoplasmic
layer, in which position the subsequent free nuclear divisions
are effected. This was the earliest stage of the endosperm ob-
served by Ikeno, who traced the subsequent stages in consider-
able detail. As the nuclei continue to divide, walls begin to
appear, and a parietal tissue is organized which gradually de-
velops into and fills up the central cavity of the embryo sac. In
this advance of tissue into the central region of the sac Ikeno
confirms the statement of Mile. Sokolowa 10 in reference to the
development of endosperm in various Conifers, that the endo-
sperm cells which are growing into the cavity of the sac have
no walls on the inner side. During the progress of endosperm
development the embryo sac continues to enlarge, encroaching
extensively upon the surrounding nucellar tissue. Warming re-
ports that in the case of Cycas circinalis, if fertilization does
not occur, the endosperm continues to grow, protrudes through
the micropyle, and being thus exposed to light develops chloro-
phyll.

The development of the archegonium has been traced in
great detail by Ikeno 28 in Cycas revoluta, and in the main con-
firms the more general accounts of previous observers. His
results are as follows: Soon after pollination (about July 1st in
Japan), two to six (mostly three, sometimes eight or more)
peripheral cells toward the micropyle become distinguishable
from their fellows by their greater size and less deeply staining
contents, and are the archegonium initials (Fig. 20, A, a\ a
condition which lasts but a few days. The second period of de-
velopment extends from the appearance of a periclinal wall in
the initial cell to the cutting off of the ventral canal cell, and
lasts over three months. By means of the periclinal wall the
initial cell is divided into an outer primary neck cell and an
inner central cell (Fig. 20, B). The former divides once by
an anticlinal wall, forming a two-celled neck, as in Giiikgo. In

CYCADALES

21

Cycas circinalis Treub J1 once or twice saw a transverse division
of one of the neck cells.

The central cell is the only remnant of the " axial row " of
the lower Archegoniates, true neck canal cells having been
eliminated from, the history of the archegonium. The great

FIG. 20. — Cycas revoluta, development of archegonium : A, upper part of endosperm,
showing an archegonium initial (a) ; B, young archegonium, showing neck cell and
central cell; (7, a portion of the layer of jacket cells surrounding the central cell,
showing the cytoplasmic connections ; Z>, two cells of the jacket layer, showing the
passage of nuclear material into the central cell; E, F, G, successive stages in the
mitosis concerned in cutting off the ventral canal cell : A-C, x 200; D-G, x about
500.— After IKENO.

enlargement of the central cell and of its nucleus, during the
three months of organization, received special attention from
Ikeno, who explains the source of the large supply of nutritive
material needed. About the central cell the adjacent endo-
sperm cells are organized into a definite jacketlike layer (Fig.

22

MORPHOLOGY OF SPERMATOPHYTES

20, (7, D). These cells have dense cytoplasm and larger nuclei,
but the very thick cellulose wall toward the central cell is
pierced by pores through which cytoplasmic threads connect
with the cytoplasm of the central cell. This continuity of cyto-
plasm between the jacket cells and the central cell in Cycads
was demonstrated by Goroschankin,12 and is confirmed by
Ikeno. In connection with his work on the embryogeny of
Cycas circinalis, Treub11-14 noted the heavy and pitted wall
bounding the central cell, and really demonstrated the conti-
nuity of the cytoplasm through it. Before the archegonium has
reached its full development the nuclei of the jacket cells show a
delicate chromatin network, but after a time they become ho-
mogeneous, showing no differentiated structure except the nucle-
olus. The contents of the jacket cells, nuclei and all, now pass
through the pores into the central cell, which thus receives an
abundant supply of nutritive material (Fig. 20, C, D). The
nucleus of the central cell, at first very small, grows during the
three months until it reaches the great size of 75 to 120/x, in
diameter. At first it has a finely granular chromatin network,
but suddenly condenses, just as the nuclei of the wall cells, and
shows no nuclear structure excepting the nucleolus. Ikeno
calls attention to the fact that this process of nutrition recalls
that of many animals, in which it has been proved that dis-
solved nutrient material passes from the follicle cells to the egg.

The third period in the preparation of the archegonium con-
sists of the formation of the ventral canal cell, which takes place
at the end of September, immediately before fertilization, and
lasts for a very short time. The central cell thrusts a beaklike
process between the neck cells, a spindle is organized at once,
and the small ventral canal cell is cut off rapidly, but for a long
time is distinguishable as a cap resting upon the ventral cell.
After the cutting off of the ventral canal cell the nucleus of the
large ventral cell passes to the central region of the organiz-
ing egg.

It is of interest to note that the occurrence of a ventral canal
cell in Cycads remained in doubt until its demonstration by
Ikeno20 in 1896 in Cycas revoluta, confirmed in 1897 by Web-
ber 27 for Zamia integrifolia. Strasburger 7 in 1876 observed
such a cell in Cycas sphaerica; Warming 8 described it in 1877
in Ceratozamia robusta, but two years later 9 withdrew his con-

(TOAD ALES

23

elusion; Treub 14 in 1884 stated positively that there is no
such cell in Cycas circinalis. Until Ikeno's demonstration in
1896, therefore, the absence of such a cell in the Cycads was
generally accepted.

It is important to note that the development of a group of
archegonia checks the growth of the vegetative tissue of the
gametophyte in that region, and the adjacent regions continu-
ing to grow, the archegonial region is left at the bottom of a pit-
like depression, which may be called the archegonial chamber,
or " endosperm cavity," as Warming has called it. This cham-
ber is full of liquid during fertilization, and at the bottom of
it the necks of the archegonia open (Fig. 18). Occasionally two
of these chambers have been observed, each with its group of
archegonia.

THE MALE GAMETOPHYTE

The male gametophyte begins with the reduction division of
the spore mother cell, which seems to have received no special
attention in Cycads. The organization of
the tetrad by successive division, accom-
panied by a peculiar chambering of the
mother cell, has been described in detail by
Juranyi5 for Ceratozamia longifolia, and
by Treub 11 for Zamia muricata (Fig. 21).
With the appearance of the nuclear plate
in the first division, a ring of thickening
develops upon the outer surface of the wall
of the mother cell in the plane of the plate.
Proceeding from this ring of thickening,
a heavy wall develops between the two
daughter cells. Juranyi thought that the
nuclear plate disappears, and that the heavy
wall is formed by growth at the free inner
edge of the ring. Treub dissents from this
view, and, although he could not demon-
strate the fact on account of the abundant
starch grains, thinks that the nuclear plate
persists and gradually thickens, the thick-
ening proceeding from the peripheral ring toward the center.
The latter view seems to be the more reasonable one. The two
3

FIG. 21. — Zamia muri-
cata ; successive stages
in the development of
the microspores : A and
j5, x 596; C, x 466.—
After TREUB.

24 MORPHOLOGY OF SPERM ATOPHYTES

daughter cells divide in the usual way, and heavy walls are
again formed. Within this heavy four-celled and often lobed
case the microspores are organized.

The first comparatively full account of the germination of
the microspore seems to have been that of Juranyi 5 for Cera-
tozamia longifolia. In 1896, in a preliminary paper,20 fol-

FIG. 22. — Cycas revoluta, development of the ciliated male cell : A, male gametophyte
of three cells, vegetative or '' prothallial " cell (/»}, generative cell, and tube cell (t\
x 500 ; B, the generative cell (a) rounded off, x 200 ; f, the generative cell (a) show-
ing division of nucleus into stalk and body nuclei, x 500 ; 7>, first appearance of
blepharoplasts in enlarging body cell (i), and diminishing stalk nucleus (s), x 500 ;
E, body cell (b) shortly before division, showing well-developed blepharoplasts (c1),
x 750 ; F, the two male cells resulting from the division of the body cell, showing
the beaked nuclei, x 200; G, II, /, later stages in the development of the ciliated
male cell, x 200. — After IKEXO.

lowed by the full paper 28 in 1898, Ikeno presented a detailed
account for Cycas revoluta; while in 1897 Webber 25> 2Q' 27 pub-
lished his results for Zamia integrifolia. One of the authors
has also examined Mr. Webber's preparations. The sequence
of events is as folloAvs (Figs. 22, 23) :

A small lenticular cell is cut off and persists, and as it seems

CYCADALES

25

to represent the vegetative tissue of the gametophyte it is often
called a " prothallial cell." There is no evidence from the lit-
erature or plates re-
ferred to above that
an evanescent vegeta-
tive cell precedes the
persistent one, al-
though the occurrence
of such a cell is possi-
ble, and it should be
looked for. At this
early stage, therefore,
the male gametophyte
consists of a single
small vegetative cell
and a large cell which
we regard as the an-
theridium initial. The
latter cell divides un-
equally, and a second
small cell is organized
which is in contact
with the vegetative
cell. This division dif- Fjo ^_Zamia integrifoiia: ± pollen tube, show.

fereiltiates the Sterile ing remnant of pollen grain (pg}, stalk cell (*),

and the fertile Series body cell (i), vegetative cell («), and tube nucleus

(£), x 80 ; B, the body cell represented in A, show-
ing the two blepharoplasts (c) surrounded by radia-
tions, x 366 ; C, division of body cell, showing the
two immature male cells and the fragmented bleph-
aroplasts (c), x 133.— After WEBBER.

A

of the antheridium.
The larger, or sterile
cell, seems to repre-
sent the antheridial
wall, but since its chief function is to develop the pollen
tube it may be called the tube cell, and its nucleus the tube
nucleus. These are the "vegetative cell" and the "vegeta-
tive nucleus " of ordinary terminology, but these phrases are
more applicable to the so-called " prothallial cell " and its
nucleus.

The smaller cell formed by the division of the antheridial
initial is the primary spermaiogenous cell, called in botanical
literature either the generative cell or " antheridial cell." Web-
ber has observed in Zamia an interesting relation between the

26 MORPHOLOGY OF SPERMATOPHYTES

vegetative cell and the generative cell, the former actively push-
ing in the wall of the latter until there is an appearance of one
cell within another. It is at this stage that Ikeno observed the
tube pushing into the nucellus and carrying with it the tube
nucleus, after which the generative cell divides to form the two
cells which have been called the stalk cell and the body cell. In
CycaSj as in Ginkgo, the plane of this division is at right angles
to the former planes of division, so that the stalk and body cells
lie side by side and are equally free to develop further. In
Zamia, however, as usually in Pinus, the plane is parallel with
the former ones, and the two cells lie fore and aft — that is, in a
lineal series with the vegetative cell. It is this position which
seems to have suggested the name stalk cell for the one next
to the vegetative cell.

The stalk cell functions no further, but the body cell enters
upon the history which is of greatest interest in the investiga-
tions of Ikeno and Webber, paralleled by Hirase for Ginkgo.
Bodies appear, named blepharoplasts by Webber,27 one at each
pole of the nucleus, their final position being in an axis at right
angles to the long axis of the pollen tube. The origin of these
bodies is unknown, but nothing to suggest them occurs in any
other cells of the plant. In their polar position they develop
remarkably in size, reaching a diameter of 10 to 15/c/, in Cycas
and 18 to 20/4 in Zamia, and giving off fibers which seem to join
with the very evident cytoplasmic network.

In the meantime the pollen tube, containing the tube nucleus,
has been growing through the nucellar tissue, and branching
more or less freely, but apparently not so profusely as in Ginkgo.
It seems to function only as an absorbing organ, for the male cells
never pass into it. A short time before fertilization the tube
nucleus passes back through the tube and consorts with the stalk
and body cells, after which the body cell divides.

This division occurs in a plane at right angles to the axis
connecting the much-enlarged blepharoplasts, and results in the
formation of two male cells, morphologically sperm mother cells,
which lie side by side, their free faces rounded, and each with a
single blepharoplast. The blepharoplast breaks up into a group
of granules, and in Cycas, according to Ikeno, a beaklike process
which remains closely associated with the blepharoplast granules
is put out from the nucleus. The elongating beak and granules

CYCADALES 27

approach the free wall and trace a widening spiral of five or
six turns, which involves about one half of the cell wall. In
this process the granules are organized into a spiral band lying
very near or against the wall. The blepharoplast band then puts
out numerous short cilia which pierce the wall of the male cell,
on whose surface a spiral groove is developed. During this
process the nuclear beak disappears, and the visible connec-
tion between nucleus and band is cut off. The final surface
appearance of the male cell is that of an oval cell, broad
and naked behind, spirally grooved in its anterior half, and
with many short motile cilia emerging from the groove. It
should be noted that Zamia? so far as Webber's preparations
show, gives no evidence of a nuclear beak associated with the
organization of the ciliated band, but it seems to move inde-
pendently toward the wall and assumes the spiral position.

During the organization of the cilia the nucleus begins to
enlarge at the expense of the cytoplasm, until finally it attains
remarkable relative dimensions, bulging between the coils of
the band, so that it becomes spirally grooved, and is covered only
by a thin mantle of cytoplasm. The ciliated band is formed
before the two sister cells separate, and Webber has observed
the pair swimming while still connected. When mature, this
ciliated cell is visible to the naked eye in Zamia, its longer axis
measuring 258 to 332/x, its shorter axis being a little less. In
Ginligo Hirase gives the dimensions as 82/4 by 49yLt; and Ikeno
reports that the dimensions in Cycas are somewhat larger, but
evidently not to be compared with those in Zamia.

It is these ciliated cells which have been called spermatozoids
or antherozoids, and such they are physiologically. Morpho-
logically, however, they are sperm mother cells which do not
organize sperms internally, but themselves pass over directly
into sperms, a fact which seems true of all Spermatophytes.
Morphologically, therefore, they are identical with the so-called
male cells of all ordinary Seed plants, but are peculiar in becom-
ing ciliated. The contrast with Pteridophytes, in which each
mother cell organizes an internal ciliated sperm and discharges
it, is sharp. The resemblance between Cycads and Pterido-
phytes is not in the morphological character of the cell which
functions as a sperm, in which the Cycads resemble the other
Seed plants, but in the fact that it is ciliated.

28 MORPHOLOGY OF SPERMATOPHYTES

It is helpful to contrast the spermatogenous series in Cycads,
which is the same as in other Gymnosperms, with that in such
a Pteridophyte as Isoetes. The sequence may be expressed
diagrammatically (Fig. 24), the terminology of the Gymno-
sperms being used.

At one time it seemed to us desirable to homologize this
series, if possible, with spermatogenesis as observed in animals,

and to use the same
terminology. Up-
on further inves-
tigation, however,
we have concluded
that such an at-
— Sperms tempt is at present

FIG. 24. — Diagram contrasting spermatogenesis in Cycads impossible In ani-
and Isoetes.

mals there is no de-
nned alternation of generations, and in their spermatogenesis
the reduction division is an important feature, so that merely
matching generation with generation in the spermatogenesis of
the two series would probably be very far from indicating any
true homologies.

It should be said that ciliated male cells have been actually
observed as yet in but three species of Cycads : in Cycas revoluta
by Ikeno, in Zamia integrifolia by Webber, and in Stangeria
paradoxa by Lang.

FERTILIZATION

Immediately after the male cells have become transformed
into ciliated sperms, the events intimately connected with fer-
tilization begin to occur. The more or less extensive pollen
tube system, which has been penetrating the loose nucellar
tissue capping the embryo sac, is abandoned, the tissue becomes
disorganized and collapses, and the gametophyte structures
within the pollen chamber are brought sensibly nearer the em-
bryo sac. The pollen tube region in connection with the pollen
grain then begins to develop, and this end of the tube, contain-
ing the tube nucleus and sperms, is turned toward the embryo
sac, its end capped by the old microspore wall (Fig. 18). Upon
reaching the wall of the embryo sac overlying the archegonial
chamber resorption of walls probably occurs, but in any event the

CYCADALES

turgid end of the tube bursts and discharges its contents into the
chamber, in whose liquid numerous sperms have been observed
swimming. As the archegonium necks open into the bottom of
this chamber, the sperms swim to them and pass down to the egg.
Webber observed in Zamia as many as four sperms in a single
neck.

As the sperm passes into the cytoplasm of the egg Ikeno and
Webber both observed that the cytoplasmic mantle with its cili-
ated band slips off, and is left in the peripheral region of the
egg cytoplasm, the sperm nucleus moving alone toward the egg
nucleus. Ikeno states that shortly before the entrance of the
sperm the egg nucleus develops a cuplike depression, into which
the sperm nucleus is re-
ceived. In any event, the
latter imbeds itself in the
former, and, according to
Webber, remains distinct
for a long time (Fig.
25). Ikeno claims that
the sperm nucleus " dis-
solves " within the egg
nucleus, and that all nu-
clear organization is lost,
but it seems more likely
that the same condition
obtains as in Conifers, a
condition which might
very easily have escaped
observation.

The prominent devel-
opment of the pollen tube
as an absorbing organ
suggests a question as to
its original significance.
Associated with ciliated
sperms, which have devel-
oped in close proximity to the egg, the function of the tube as
a sperm carrier is reduced to a minimum, and in fact does not
seem to occur in the Cycads. It is an interesting question,
therefore, whether the pollen tube originated as an absorbing

FIG. 25. — Fertilization in Cycads : A, fusion of
the nuclei in Cycas revoluta, t indicating the
so-called trophoplasmic substance ; .Z?, fusion
of the nuclei in Zamia integrifolia, b being
the discarded ciliated band, x 20; 0, the male
nucleus (mn) of Zamia integrifolia after slip-
ping from the cytoplasmic sheath and ciliated
band, x 66.— A, after IKENO ; B and C, after
WKBBER.

30

MORPHOLOGY OF SPERMATOPHYTES

organ and later became used as a sperm carrier, or whether
the reverse is true. It would seem natural to think of the
Cycad condition as the more primitive one.

IV. THE EMBRYO

The first comparatively full account of the development of
the embryo is that given by Treub 14 for Cycas circinalis, and
this account has been confirmed by fragmentary observations
The account given by Ikeno 28 for Cycas revoluta agrees

since.

FIG. 26. — Cycas circinalis, development of the embryo: A, an ovule from above, show-
ing the openings of the necks of several archegonia, x % ; B, longitudinal section
of'an archegonium, x 73; (7, a fertilized egg, showing several free nuclei, x 24;
D, free nuclei of the proembryo, showing parietal placing, x 36 ; E, a young pro-
embryo, x 15; F. a young embryo, showing saclike proembryo, suspensor (susp),
and embryo proper («), x 6; G, longitudinal section of seed containing two embryos,
two thirds natural size; If, embryo with suspensor, x 1%; </, a more advanced em-
bryo, two thirds natural size ; K, a mature embryo. — After TREUB (from Engler and
Prantl's Nat. Pflanzenfam.).

in every particular with Treub's, and fills in certain gaps.
There is still almost entirely lacking any knowledge of the devel-
opment of the embryo itself. The embryogeny shows three dis-
tinct phases as follows (Fig. 26) :

1. Development of Proembryo. — This name was applied
by Treub to the intrasporic development. The oospore en-
larges, continuing to receive supplies of nutrition from the adja-
cent tissue. The first change observed by Treub was the pres-

CYC A DALES 31

ence of numerous free nuclei distributed through the cytoplasm.
He naturally inferred that these were derived by repeated
nuclear division from the oospore nucleus. Ikeno observed the
first division of this nucleus, and has called attention to certain
peculiarities connected with it — peculiarities which are chiefly
suggestive in being found also in Ginkgo and in Pinus. The
first nuclear spindle is the largest one which appears during free
nuclear division, and its axis is always inclined to the long axis
of the archegonium. Whether this position is significant or
not remains to be seen, but the fact that it is also observed in
Ginkgo and Pinus suggests that it is worthy of record. The
fibers of this primary spindle do not converge as they do in all
subsequent ones. As a consequence, this spindle is very broad
in outline, really rectangular, while all subsequent spindles are
distinctly bipolar. A further point of interest in connection
with this spindle is that there are no centrosomes, or centrosome-
like bodies, or even polar radiations connected with it. It would
seem, therefore, that there is no organ which might be regarded
as the lineal descendant of the blepharoplast, which was so promi-
nent in the sperm series. The subsequent bipolar spindles show
the ordinary radiations, but no bodies at the poles.

Free nuclear division proceeds with rapidity and in simul-
taneous fashion until the nuclei become very abundant. Ikeno
xcalls attention to the fact that at this time the structure of the
cytoplasm of the central region changes. It becomes at first vac-
uolate, and then completely disorganizes, the nuclei imbedded in
it sharing the same fate. The remaining cytoplasm becomes
parietal, massing somewhat towards the bottom of the spore. In
this parietal layer the nuclei are imbedded, being equidistant
from one another and forming usually a single layer, but at the
base they are more massed. In this position the nuclei con-
tinue to divide and reach large numbers, when cell walls begin
to appear. An interesting fact in connection with this division
of nuclei, when in the parietal position, is that some of the free
nuclei which are nearest the neck end of the oospore divide ami-
totically, but this tendency seldom reaches into the deeper re-
gions. As a result of this development the proembryo becomes
a sac, somewhat thickened at the base, but with the wall com-
posed of one or at most two layers of cells. At this point the
work of Ikeno stops, and beyond it Treub gives very little detail.

32 MORPHOLOGY OF SPERMATOPHYTES

2. Development of Suspensor. — At the beginning of the sec-
ond period of development the cavity of the saclike proembryo
fills up with tissue somewhat at the base, but is never obliter-
ated. This second period, however, is chiefly devoted to the
development of the suspensor, which is effected by the elonga-
tion of cells in the mass at the base of the spore which lie imme-
diately behind the terminal group. These cells, by elongating
and dividing, develop a remarkably long and tortuous and mass-
ive suspensor, which has never been observed to branch or to
give rise to more than a single embryo. The suspensors emerge
from the archegonium region in a fascicle, and, imbedded in
the endosperm, form a considerable cavity by the disorgan-
ization of the adjacent cells. The tips of the suspensors bear-
ing the groups of cells destined to form the embryos are im-
bedded in the tissues beyond this cavity. The suspensor is
quite persistent, remaining attached to the mature embryo,
and suspensor coils with abortive embryos may be seen even
in ripe seeds.

3. Development of Embryo. — Nothing of morphological sig-
nificance is known in reference to the development of the em-
bryo. The earliest stages figured show an embryo far advanced.
In its complete state the embryo shows two cotyledons, but con-
cerning the origin of these cotyledons, or of the stem tip, or of
the root tip, we have no information.

THE SEED

The seed becomes plumlike, the testa developing from the
thick integument, in two layers, the outer fleshy, the inner
stony. The view that these two very distinct layers of the testa
may represent two integuments, which have been merged into
what appears to be a single very thick integument, will be re-
ferred to later, in connection with the other groups. In Cycas
revoluta, and probably in the other Cycads, the development of
the seed does not depend upon fertilization, for with or without
pollination there is the same increase in size and development
of the two layers of the testa. In germination the cotyledons
are said to be hypogean and closely coherent at the apex, as in
j but no definite work in seed germination among Cycads
seems to have been done.

CYCADALES 33

LITERATURE CITED

1. BRONGNIART, A. Recherches sur 1'organisation des tiges des Cy-

cadees. Ann. Sci. Nat. I. 16: 389-402. pis. 20-22. 1829.

2. VON MOHL, HUGO. Ueber den Bau des Cycadeen-stammes. Abh.

der K. b. Acad. zu Miinchen 1: — . 1832. Republished and
revised in Vermischte Schrifteu 195-211. 1845.

3. MIQUEL, F. A. W. Monographia Cycadearum. 1841.

4. METTENIUS, G. H. Beitrage zur Anatomie der Cycadeen. Abh.

der K. Sachs. Gesells. der wiss. 7 : 565-608. pis. 1-5. 1861.

5. JURANYI, L. Bau und Entwick. des Pollens bei Ceratozamia

longifolia Miq. Jahrb. f . wiss. Bot. 8 : 382-400. pis. 31-34. 1872.

6. REINKE, J. Parasitische Anabaena in Wurzelen der Cycadeen.

Gott. Nachr. 107. 1872.

7. STRASBURGER, E. Ueber Zellbildung und Zelltheilung. 1876.

8. WARMING, E. Recherches et remarques sur les Cycadees. Over-

sigter over d. K. D. Vidensk. Selsk. Forh. 1877.

9. WARMING, E. Contributions a Thistoire naturelle des Cycadees.

Ibid. 1879.

10. SOKOLOWA, MLLE. C. Naissance de 1'endosperme dans le sac

embryonnaire de quelques Gymnospermes. Moscou. 1880.

11. TREUB, M. Recherches sur les Cycadees. Ann. Jard. Bot. Buit. 2 :

32-53. pis. 1-7. 1881 (date of reprint, volume date being 1885).

12. GOROSCHANKIN, J. Zur Kenntniss der Corpuscula bei den Gym-

nosperinen. Bot. Zeit. 41 : 825-831. pi. 7 A. 1883.

13. DE BARY, A. Comparative Anatomy of the Vegetative Organs of

the Phanerogams and Ferns. English translation. 1884.

14. TREUB, M. Recherches sur les Cycadees. 3. Embryogenie du

Cycas circinalis. Ann. Jard. Bot. Buit. 4 : 1-11. pis. 1-3. 1884.

15. COSTANTIN and MOROT. Sur Torigine des faisceaux libero-ligneux

supernumeraires dans la tige des Cycadees. Bull. Soc. Bot.
France 32: 173. 1885. Ref. Bot. Centralbl. 24: 101-102. 1885.

16. GOEBEL, K. Outlines of Classification and Special Morphology.

English translation. 316. 1887.

17. SOLMS-LAUBACH, H. GRAF zu. Die Sprossfolge der Stangeria und

der iibrigen Cycadeen. Bot. Zeit. 48 : 177-187, 193-199, 209-215,
225-230. 1890.

18. STRASBURGER, E. Histologische Beitrage 3 : 151. 1891.

19. SCHNEIDER, A. Mutualistic Symbiosis of Algae and Bacteria with

Cycas revoluta. Bot. Gaz. 19 : 25-32. pi. 3. 1894.

20. IKENO, S. Das Spermatozoid von Cycas revoluta. Bot. !Mag.

Tokyo 10: 367-368. 1896.

21. WORSDELL, W. C. Anatomy of Stems of Macrozamia compared

with that of other Genera of Cycadeae. Ann. Bot. 10: 601-620.
pis. 27-28. 1896.

34 MORPHOLOGY OF SPERMATOPHYTES

22. IKENO, S. Vorlaufige Mittheilung iiber die Spermatozoiden bei

Cycas revoluta. Bot. Centralbl. 69 : 1-3. 1897.

23. LANG, W. H. Studies in the Development and Morphology of

Cycadeaii Sporangia. I. The Microsporangia of Stangeria
paradoxa. Ann. Bot. 11: 421-438. pi 22. 1897.

24. SCOTT, D. H. The Anatomical Characters presented by the Pedun-

cles of Cycadaceae. Ann. Bot. 11 : 399-419. pis. 20-21. 1897.

25. WEBBER, H. J. Peculiar Structures occurring in the Pollen Tube

of Zamia. Bot. Gaz. 23 : 453-459. pi. 40. 1897.

26. WEBBER, H. J. The Development of the Antherozoids of Zamia.

Bot. Gaz. 24: 16-22. 1897.

27. WEBBER, H. J. Notes on the Fecundation of Zamia and the Pol-

len Tube Apparatus of Ginkgo. Bot. Gaz. 24: 225-235. pi. 10.
1897.

28. IKENO, S. Untersuchungen iiber die Entwickelung der Gesch-

lechtsorgane und der Vorgang der Befruchtung bei Cycas revo-
luta. Jahrb. f. wiss. Bot. 32 : 557-602. pis. 8-10. 1898.

29. LANG, W. H. Studies in the Development and Morphology of

Cycadean Sporangia. II. The Ovule of Stangeria paradoxa.
Ann. Bot. 14: 281-306. pis. 17-18. 1900.

30. WORSDELL, W. C. The Anatomical Structure of Bowenia specta-

bilis. Ann. Bot. 14: 159-160. 1900.

31. LIFE, A. C. The Tuberclelike Boots of Cycas revoluta. To appear

in Bot. Gaz. 31 : F 1901.

CHAPTER II

GINKGOALES

THIS group is represented by a single living species, Ginkgo
biloba, almost unknown in the wild state, but reported in certain
forests of western China. Its extensive cultivation, however,
first in China and Japan, and later in all civilized countries, has
made it accessible. It was long regarded as an exceptional
Conifer of the Taxus group, although its resemblances to Cycads
were frequently pointed out. Recent investigation, however,
has resulted in regarding this species as the only survivor of an
ancient phylum which deserves to be coordinated with Cycads
and Conifers. A very full general account of the species has
been published recently by A. C. Seward and Miss J. Gowan,16
while two papers by Hirase llf 13 supply the first comparatively
full account of the reproductive structures.

I. THE VEGETATIVE ORGANS

THE STEM

The tree has the general habit of a Conifer, with central
shaft and widespreading branches (Fig. 27). It is recorded as
reaching sometimes a height of more than 30 meters, with a
circumference of more than 8 meters. The vascular bundles are
collateral, and with a persistent primary cambium as in the Coni-
fers, resulting in growth rings of the ordinary type. The sec-
ondary wood is composed of tracheids with bordered pits, and a
characteristic double leaf trace connects the vascular system
of the stem with that of theleaves. Resin ducts are abundant,
both in the pith and in the cortex. Two types of branches may
be distinguished, the long shoots which are terminal and elon-
gate rapidly, and the short shoots which are axillary and slow
growing, usually bearing tufts of leaves at the summit, and
covered below with old leaf scars.

35

36

MORPHOLOGY OF SPERMATOPHYTES

THE LEAF

The foliage leaves are very characteristic in form and vena-
tion, resembling in this regard some of the Ferns. The petiole

i

•Giiikgo biloba. Taken originally from a water-color painting by a Chinese
artist. — After SEWAKD and Go WAN.

is long and slender, the blade is broadly wedge-shaped and vari-
ously lobed, and the venation is distinctly dichotomous (Figs.

GIXKGOALES

37

28, 29). There is great variability, however, both in size and

amount of lobing.

The development of the foliage leaf has been described by

Fankhauser.6 It is at first a protuberance embracing two fifths

of the circumference at the stem

apex. Soon a distinct emargina-

tion is developed which becomes

a deep incision. During devel-

opment the blade is bent over at

the apex and the margins are

strongly inrolled. The leaves,

especially when young, are not

merely deeply two-lobed, but

each lobe is variously toothed

or divided.

In anatomical characters the

leaf resembles those of Cycads.

There is developed an evident

but not very thick cuticle. The

stomata are restricted to the

lower surface, the guard cells

being somewhat below the level

of the epidermis. A palisade

tissue is evident, but somewhat irregular. A transfusion tissue,

such as was described under the Cycads, appears also in the

leaves of Gink-
go, cells of the
inner mesophyll
elongating in a
plane parallel
with the leaf
surface and de-
veloping numer-
ous and promi-
nent intercellu-
lar spaces.

I M f*

FIG. 28. — Ginhgo biloba, a cluster of
leaves and a mature seed.

FIG. ^.— Ginl-no biloba, a single leaf, showing venation.

are deciduous, a

rare habit among Gymnosperms, being displayed in Conifers
only by Larix, Taxodium disticlium, and Glyptostrobus.

38 MORPPIOLOGY OF SPERMATOPHYTES

THE BOOT

Some account of the anatomy of the root has been published
by Van Tieghem.2 It seems that the vascular cylinder is origi-
nally a diarch, and a little later each strand divides, resulting
in the formation of a tetrarch cylinder. Later, new strands
come in, so that five or six alternating xylem and phloem bun-
dles appear. Presumably the meristem of the apex shows the
same peculiarities as described under Conifers.

II. THE SPORE-PRODUCING MEMBERS

THE MICROSPORANGIUM

The stamens occur in loose catkinlike clusters from the axils
of the scale leaves developed at the summit of the short shoot
(Fig. 30, B). It is this elongated catkinlike cluster of micro-
sporophylls which has received the inappropriate name of the
" male flower." The sporophyll has a rather long stalk and
terminates in a more or less expanded knob, which may be re-
garded as representing the expanded sterile tip common among
the Cycads. Beneath the knob the pendent sporangia are borne,
usually two in number, but sometimes three, or even four. The
dehiscence is by a longitudinal slit (Fig. 30, (7).

There are no published details of the development of the
sporangium, but there is no reason to suppose that it is different
in any important way from that described for the Cycads. The
mature sporangium wall consists of four to seven layers of cells,
with thickening bands on the outer layers.

THE MEGASPORANGIUM

The structures associated with the megasporangium have
given rise to much discussion. A long stalk arises from an axil
of the leaves borne at the summit of a short shoot, and near its
apex mostly two megasporangia are borne, only one of which
functions (Fig. 30, A, E). About the base of each sporangium
is a more or less conspicuous cup or " collar," much more prom-
inent in the younger stages of the sporangium, reaching its full
development very early. Seward and Gowan 15 have published
a resume, of the various views held in reference to the homolo-
gies of these structures. Briefly they are as follows:

GIXKGOALES

39

In 1809 Van Tieghem * published his view that the stalk
represents a petiole, and that the two ovules are determined by
the two characteristic lobes of the blade. According to this
view, the whole structure stands for a single carpel. Van
Tieghem further sees in the collar at the base of each ovule a
reduced arillus, which name of course gives no clew to its
homology. In 1872 Strasburger 3 published the view that
the stalk is a shoot, and that the collar is the mdiment of the
first pair of leaves of a secondary shoot. It would follow that
the whole structure is an inflorescence bearing two flowers, and

FIG. ZQ.—Giitkyo biloba: A, dwarf branch bearing leaves and young ovules; J?, a stam-
inate strobilus; 61, more enlarged view of a portion of B; Z>, longitudinal section of
a young ovule ; E, a ripe seed associated with an undeveloped ovule. A, B, E,
natural size ; C*, Z>, enlarged. — After GOEBEL.

that any such structure as a carpel is suppressed. In 1879 5
he modified his view as to the character of the collar, having
concluded, as Van Tieghem had thought, that it represents an
arillus. He cited cases in which a stalk bore four ovules, each
upon a slender stalk of its own; this of course was a strong
argument in favor of the view that the whole structure repre-
sents a shoot. In 1873 Eichler 4 published his view that the
collar represents an outer integument, which is not necessarily
inconsistent with the preceding views that it stands for an
arillus. Later 7 he called the collar a rudimentary carpel, and
4

MORPHOLOGY OF SPERM ATOPHYTES

regarded the two ovules as representing a single flower. In
1890 Celakovsky 8 published the view that the stalk is an axil-
lary shoot bearing two or more carpels, but that each carpel is
represented only by an ovule. This means that the presence of
the carpel is purely theoretical. The same author 14 in 1900
reaffirms this position with greater fullness of detail. In 1896
Fujii,12 after a study of abundant Japanese material, and espe-
cially taking into account numerous abnormal developments,
came to the conclusion that the stalk is a shoot, usually bearing

two rudimentary carpels. That
the ovules are really related
to carpels appears from the
fact that he found them upon
more or less modified foliage
leaves. In other cases he
found transition stages be-
tween the normal collar and
blades bearing ovules, his con-
clusion being that the collar
represents the rudimentary
carpel. The short stalk which
is sometimes quite evident in
connection with each ovule
stands, therefore, for the peti-
ole of a megasporophyll. That
the main stalk is a shoot was
further evidenced by the fact
that cases were found in which
a stalk bore several ovules and
terminated in a scaly bud. The
results of Seward and Gowan's 16 studies coincide with the views
of Fujii. These observers have still further strengthened the
position by the record of additional abnormalities.

A summary of what seems to be the most reasonable view of
I the homologies of the ovulate structures in Ginkgo is as follows :
j The main stalk represents a shoot bearing two or more carpels,
] and these carpels mostly occur in rudimentary form, the petiole
I being represented by the short stalk upon which each ovule
Stands, and the blade by the so-called collar.

The development of the megasporangium seems to be al-

FIG. 31. — Ginkgo Mloba, longitudinal sec-
tion of ovule : i, integument ; m, mi-
cropyle; p, beaked apex of nucellus,
with pollen chamber containing pollen
grains; <?, embryo sac containing an
early stage of the gametophyte ; c, the
"collar"; x 5.

GIXKGOALES 41

most identical with that in the Cycads. According to Pax,9
a many-celled archesporium is differentiated. The single thick
integument with a long and narrow micropyle, the large mass of
sterile nucellar tissue above the single megaspore, the conspicu-
ous and resistant nucellar beak projecting into the micropyle
and containing a pollen chamber, the comparatively loose nu-
cellar tissue between the beak and the megaspore, are all as in
the Cycads (Fig. 31). Hirase 13 says that the pollen chamber is
organized early in May, soon after the maturing of the pollen,
and that near the time of pollination it is full of liquid. He
also observes that the chamber results from the exclusive devel-
opment of the external tissue of the beak, the inner tissue thus
becoming ruptured and disorganized. The changes which occur
in the nucellus in connection with the development of the pollen
tubes will be described under fertilization.

III. THE GAMETOPHYTES

THE FEMALE GAMETOPHYTE

The development of the endosperm seems to proceed in the
same way as in the Cycads, although such full details* are
not on record. However, the following sequence of events
is the same: free nuclear division, parietal placing of free
nuclei imbedded in a cytoplasmic layer, appearance of walls
resulting in the organization of a tissue, and gradual growth of
the parietal tissue toward the center of the sac, which it finally
completely fills. During this development the sac encroaches
upon the adjacent nucellar tissue, practically obliterating all of
it except that which caps the sac. Hirase has made the inter-
esting observation that about 256 free nuclei have been organ-
ized when walls begin to appear, representing eight successive
nuclear divisions.

There are usually two archegonia, but sometimes a greater
number has been observed. They are exactly as in the Cycads,
with two neck cells, and a central cell which develops remark-
ably in size and nutrient material (Fig. 34, A-C). As in Cy-
cads, also, the adjacent cells of the endosperm organize a layer
of jacket cells about the central cell, and a heavy wall is de-
veloped. It seems very probable that the central cell in this
case receives nutritive supplies from the jacket cells through

42 MORPHOLOGY OF SPERMATOPHYTES

wall pores, as has been demonstrated in Cycads and Conifers.
Just before fertilization, early in September, the small and
ephemeral ventral canal cell is cut off, and the large egg is organ-
ized (Fig. 34, (7). Hirase states that about twenty weeks elapse
between pollination and the maturity of the egg. It should
also be noted that an archegonial chamber is developed in the
endosperm, as in Cycads, though this is somewhat modified later
by the remarkable development of an endosperm beak, to be
described in connection with fertilization.

THE MALE GAMETOPHYTE

Our knowledge of the development of the male gametophyte
of Gink go dates from Strasburger's account 10 in 1892, much
extended by the recent very complete observations of Hirase 13
(Fig. 32). From this last account the following description of
the male gametophyte and of fertilization is derived.

The first division of the microspore is unequal, resulting in
cutting off a small lenticular cell, which soon disorganizes and
later appears merely as a cleft in the heavy spore wall. This
first ephemeral cell may be regarded as a vegetative cell of the
gametophyte, a vestige of the vegetative tissue of a more exten-
sive and probably independent gametophyte. It is interesting to
note that this same ephemeral cell appears in the germination
of the microspores of most of the Conifers, but was not observed
by either Ikeno or Webber in the Cycads.

A second unequal division occurs, and a second small cell
is cut off, which persists, and which represents a second vegeta-
tive cell (Fig. 32, A, p). This second vegetative cell is cut
off in the microspore of most Conifers, but it is ephemeral like
the first; while this single persistent vegetative cell appears to
be the first and only one cut off in the microspores of Cycads.
Ginkgo resembles the Conifers, therefore, in having two vege-
tative cells ; and resembles the Cycads in having one persistent
vegetative cell. It is a question whether a first ephemeral vege-
tative cell may not yet be detected in Cycads, thus confirming
the close resemblance between these two groups which appears in
almost every other feature of their male gametophytes.

After the two vegetative cells have been cut off, the remain-
ing large cell is what we regard as the antheridium initial, whose
function is to organize the single antheridium. This cell di-

GTXKGOALES

vides unequally, the smaller daughter cell, in contact with the
persistent vegetative cell, being the primary spermatogenous
cell or generative cell (Fig. 32, B, a), the larger daughter cell
being the tube cell (Fig. 32, J5, t). The gametophyte is in this

C

FIG. 32. — GinTcgo biloba, development of male gametophyte: A, pollen grain collected
April 24th, showing third nuclear division and the persistent vegetative cell (p),
•x. 500; B, a later stage, showing persistent vegetative cell (/>), generative cell (a),
and tube nucleus (2), x 500; f, later stage, July 10th, the generative cell almost
spherical, x 500; D, division of nucleus of generative cell into nuclei of stalk and
body cells, July llth, x 500; F, division of the body cell, the blepharoplasts being
faintly visible, x 120; E, the two male cells produced by the body cell, September
12th. x 22rt.— After HIRASE.

three-celled stage when pollination takes place, during the last
of April or early in May.

From this stage on the sequence is exactly as that described
for Cycas. The generative cell divides to form the stalk and
body cells (Fig. 32, D), which lie side by side, only the body
cell functioning further. In the body cell the two blepharo-
plasts (Fig. 32, F) appear, and division occurs, resulting in
the two male cells, lying side by side, and each containing
a blepharoplast. Each male cell then directly develops as a

44: MORPHOLOGY OF SPERMATOPHYTES

multiciliate sperm, the blepharoplast forming the cilia-bearing
band, which a beaklike process put out by the nucleus seems to
guide in organizing the spiral of about three coils.

Although this general account is the same as that given for
Cycads, there are at least two differences in detail which deserve
mention. Hirase states that the division of the generative cell
of Ginkgo is only a nuclear division, or at least that there is no
dividing wall, and that the nonfunctional nucleus is gradually
thrust out of the cell by the large development of the function-
al nucleus. He further states that when the cilia-bearing band
is about to be organized by the blepharoplast, the latter puts out
a hooklike process which attaches itself to the nucleus, and fol-
lowing this the nuclear beak develops with the hooklike process
attached to its tip. Hirase's observations indicate that the
initiative which results in the nuclear beak proceeds from
the blepharoplast, while Ikeno's observation indicates that the
initiative proceeds from the nucleus itself. It will be remem-
bered that Webber observed no nuclear beak in Zamia.

The homology of these ciliated cells functioning as sperms,
and the reason for not regarding them as the morphological
equivalents of the multiciliate sperms of Pteridophytes, but
rather the equivalents of the so-called male cells common to all
Spermatophytes, are fully stated in connection with the Cy-
cads. Hirase's figures show a wall inclosing the ciliated cell,
presumably the wall of the mother cell (male cell), but no
reference is made to it in the text. If this represents the real
situation, it would follow that the ciliated cell in Ginkgo is mor-
phologically a sperm, being organized within a mother cell and
discharged from it. This view is clearly supported by the re-
cent figures and statement of Fujii,14 \vhich represent the multi-
ciliate sperm as distinctly organized within the mother cell and
discharged from it. The details of development are so nearly
identical with those in Cycas and Zamia, that it seems improb-
able that such an important difference could occur. If it be true
that the ciliated sperm in Ginkgo is morphologically a sperm, and
not merely a ciliated mother cell, it would follow that Ginkgo is
still more fernlike than was supposed, and unlike other Gymno-
sperms and all Angiosperms in the feature that its sperm mother
cells do not function as male cells. It must be remembered that
Ginkgo is a much older type than are the Cycads investigated,

GIXKGOALES

45

even the genus Cycas, and it is not inconceivable that this differ-
ence between a very ancient type and much more modern types
may occur.

FERTILIZATION

We have reserved for this topic all of the events associated
with the development of the pollen tubes, as well as the imme-
diate act of fertilization. When the pollen grains reach the
pollen chamber, about the first of May, the male gametophyte is
in its three-celled stage, the cells being the persistent vegetative
cell, the generative cell, and the large tube cell. The pollen
chamber now begins to enlarge at the expense of the nucellar
tissue beneath, carrying the grains deep into the nucellus. By
the beginning of June the chamber has greatly enlarged, form-
ing a large irregular cavity, which practically reaches the
embryo sac. When the grains begin to put out their tubes, the
opening to the pollen chamber becomes closed and remains so,
the surrounding tissue becoming brown and forming a solid
beak which long persists as a cap on the sac. The deep placed
grains, carried by the enlarging chamber well below the beak,
send out their tubes in every direction into the adjacent nucellar
tissue, but chiefly away from the embryo sac, and often directly
toward the beak. Into these young tubes the tube nuclei pass,
and remain there as long as the tube system is developing.

About the beginning of July the tubes have penetrated the
nucellar tissue more deeply and branch freely, the tube nucleus
remaining where a tube begins to branch. This extensive sys-
tem of branching tubes seems to function only as an absorbing
system, as in the Cycads. At this stage the generative cell has
the appearance of being pushed away from contact with the
vegetative cell, on account of the development of vacuoles in the
latter (Fig. 32, D); it increases very much in size, its nucleus
keeping pace with it ; and these relative positions are main-
tained until just before fertilization, which occurs about ten
weeks later.

About the middle of July the nucleus of the generative cell
divides rapidly, forming two daughter nuclei which represent
the stalk and body cells. One of these nuclei enlarges very
much, and comes to occupy a central position in the cell, crowd-
ing the other aside into the space between the generative cell
and the vegetative cell. In this way the body cell is organized.

MORPHOLOGY OF SPERM ATOPIIYTES

At the end of July the body cell increases still more in size,
becoming rich in contents, and its nucleus becoming very large.
At this stage the blepharoplasts appear, and finally assume their
polar positions at the extremities of the long axis of the ellip-
soidal nucleus. This stage persists for three weeks — that is,
until the third week before fertilization.

At the beginning of August the tube nucleus begins to pass
back to the grain end of the tube, which it reaches in about two
weeks, consorting with the body cell, or with the male cells,
until fertilization. During this retreat of the tube nucleus and
abandonment of the absorbing tube system, the endosperm be-
gins to develop a beak between the two archegonia. Hirase sug-
gests that its significance is to be explained by the fact that the

tissue between the nucellar
beak and the embryo sac has
been broken down by the
deepening of the pollen cham-
ber and by the absorptive work
of the pollen-tube system, and
that the heavy beak settles
down upon the sac. In about
two weeks the endosperm beak
looks like a small column,
with its summit against the
nucellar beak, " like a tent
supported by its center pole,"
in the shelter of which there is freedom for the final processes
connected with fertilization. Pressure upon the grain ends of
the pollen tubes which happen to lie under " the tent " is thus
avoided, and these ends become very turgid and are directed
toward the near-lying archegonia (Fig. 33).

During the third week before fertilization — that is, about
the last of August — the body cell begins division, and the cili-
ated male cells are organized. The swollen tip of the pollen
tube, capped by the old wall of the pollen grain, now contains
in a group the two male cells, the tube nucleus, and the vege-
tative cell, together with whatever may remain of the stalk
cell. At the time of fertilization the capped tip of the tube
turns toward the archegonial chamber and discharges the con-
tents into it. The exact position of the sac wall in relation to

FIG. 33. — Ginlcgo Mloba, upper part of fe-
male gametophyte, showing two arche-
gonia, the beaklike process of the en-
dosperm supporting the remains of the
nucellus, and the swollen tips of two
pollen tubes, September 9th, x 24. — After
HIRASE.

GINKGOALES 47

the archegonial chamber after the development of the endo-
sperm beak is not made clear in Hirase's account, which seems
to imply that the way is open to the archegonial chamber. In
any event, the sperms reach the chamber, and in it Hirase has

n

D

FIG. 34. — Ginlcgo biloba: A, longitudinal section of the neck of the archegonium, x 160;
B, transverse section of the same, x 160 ; Ct archegonium shortly before fertilization,
showing ventral canal cell (v) and egg nucleus («), x 66 ; Z>, free nuclei resulting
from the first nuclear divisions of the oospore, x 66 ; E^ later stage, showing the
compact tissue of the embryo tilling the oospore, x 66. — After STRASBURGEK.

observed them swimming. The entrance of sperms into the
archegonia follows, and fertilization is effected.

The statement has been usual that fertilization does not
occur until the ripe ovules have fallen. While this may be
true frequently, Hirase found that some seeds, at least, con-

48

MORPHOLOGY OF SPERMATOPHYTES

tain embryos while still attached, a fact which we have been
able to verify in our own experience.

IV. THE EMBRYO

Although the embryo of Ginkgo is exceptional among Gym-
nosperms, and of great interest, the details of its development

FIG. 35.— Photographs of embryos of Ginkgo biloba in various stages: A, young
embryo; B, older stage, showing cotyledons ; (7, a still older embryo, showing sev-
eral young leaves ; Z>, a portion of a nearly mature embryo, with conspicuous resin
ducts in stem, cotyledons, and leaves ; the beaklike process supporting the remains
of the nucellus is still visible.

GINKGOALES

49

are not sufficiently known. We have been able to secure almost
a complete series showing the general outlines of the develop-
ment (Figs. 34-36), which merely confirms the facts already
published. Germination of the oospore begins, as is usual,
among Gymnosperms, with repeated nuclear division (Fig.
34, D). These nuclei, however, instead of organizing a parietal
tissue as in the Cycads, or a basal group as in the Conifers,
proceed to fill the whole cavity of the enlarging oospore with
free nuclei, which is followed by the organization of a com-
pact tissue (Fig. 34, E). In a certain sense this structure would
seem to represent the proembryo of Cy-
cads, but it really represents the whole
product of the oospore, in which proem-
bryo, suspensor, and embryo proper are not
differentiated. The complete filling of the
spore with tissue, and the lack of early
differentiation into the great embryonic
regions, would suggest a habit more primi-
tive than in either Cycads or Conifers. At
the same time, it may be merely a derived
character. In any event, the tissue near
the base of the spore, which in the other
groups develops both suspensor and embryo,
shows far greater vigor than the remain-
ing tissue. In the organization of the em-
bryo the whole mass of tissue is involved, and in the absence
of a suspensor the embryo invades the endosperm by direct
growth.

The two cotyledons are differentiated early in October, and
are quite unequal in length. The larger one is two-lobed, while
the shorter one is cleft halfway down, thus early indicating the
bilobed character of the leaf. The two cotyledons are also
united at the apex, but the epidermal layers of the two are
distinct where they are in contact. The plumule is very con-
spicuous between the elongated cotyledons, three or more leaves
appearing just behind the stem apex (Fig. 35).

As in the Cycads, the seed becomes plumlike, a testa with
fleshy outer and stony inner layers being organized (Fig. 36).
Even without pollination the seed attains its usual size and the
two layers of the testa are developed.

FIG. Zb.— Ginkgo biloba,
a seed with embryo in
stage shown in Fig.
35, C: e, embryo; g,
endosperm ; k and /,
hard and fleshy layers
of the testa ; natural

50 MORPHOLOGY OF SPERMATOPHYTES

LITERATURE CITED

1. VAN TIEGHEM, PH. Aiiatomie comparee de la fleur femelle et du

fruit des Cycadees, des Coniferes, et des Gnetacees. Ann. Sci.
Nat. V. 10: 269-304. pis. 13-16. 1869.

2. VAN TIEGHEM, PH. Recherches sur la symetrie de structure des

plantes vasculaires. Ann. Sci. Nat. V. 13 : 1-314. pis. 3-8. 1870.

3. STRASBURGER, E. Die Coniferen und die Giietaceen. 1872.

4. EICHLER, A. W. Sind die Coniferen gymnosperm oder nicht ?

Flora 56: 241-247, 260-272. 1873.

5. STRASBURGER, E. Die Angiospermen und die Gymnospermen.

1879.

6. FANKHAUSER, J. Entwickeluiig des Stengels und Blattes von

Ginkgo. Bern, 1882.

7. EICHLER, A. W. Coniferae in Engler and Prantl's Die nat. Pflanz.
^ 21: 108. 1889.

8. CELAKOVSKY, L. Die Gymnospermen Abh. k. bohm. Gesell.

wiss. VII. 4: 1 seq. 1890.

9. PAX, F. Allgemeine Morphologic der Pflanzen mit besonderer

Beriicksichtigung der Bliithenmorphologie. 1890.

10. STRASBURGER, E. Ueber das Verbal ten des Pollens und die

Befruchtungsvorgange bei den Gymnospermen. 1892.

11. HIRASE, S. Etudes sur la fecondation et 1'embryogenie du Ginkgo

biloba. Jour. Coll. Sci. Imp. Univ. Tokyo 8: 307-322. pis.
31-32. 1895.

12. FUJII, K. On the Different Views hitherto proposed regarding

the Morphology of the Flowers of Ginkgo biloba. Bot. Mag.
Tokyo 10: 15-25, 104-110. pi. 1. 1896.

13. HIRASE, S. Etudes sur la fecondation et 1'embryogenie du Ginkgo

biloba. Jour. Coll. Sci. Imp. Univ. Tokyo 12 : 103-149. pis. 7-9.
v 1898.

14. CELAKOVSKY, L. Die Vermehrung der Sporarigien von Ginkgo

biloba L. Oesterreich. hot. Zeitsch. 50: 229-237, 276-283, 337-
341. 1900.

15. FUJII, K. On the Morphology of the Spermatozoid of Ginkgo

biloba. Bot. Mag. Tokyo 14: 260-266. pi. 7. 1900.

16. SEWARD, A. C., and GOWAN, Miss J. The Maiden Hair Tree

(Ginkgo biloba L.). Ann. Bot. 14 : 109-154. pis. 8-10. 1900.

CHAPTER III

CONTFERALES

THE representative Gymnosperms, at least of the present
time, are Conifers. They include about three hundred recog-
nized living species, distributed among thirty-four genera, and
constitute fully three fourths of the present Gymnosperni flora.
They are characteristic of the temperate regions of both north-
ern and southern hemispheres, presenting a sharp contrast in
this regard to the Cycads.

I. THE VEGETATIVE ORGANS

THE STEM

The most conspicuous morphological feature of the stem is
its regular monopodial branching. The terminal bud continues
to be more vigorous than the lateral ones, both in the main axis
and in the branches, and as a result the characteristic symmetri-
cally conical form of the Conifers is developed (Figs. 37—40).

The branches are distinctly dimorphic, the two forms being
characterized as long shoots and spur shoots. In what may be
regarded as the more primitive forms of the stem body both
kinds of shoots bear foliage leaves ; but in Pitius the main axes
bear only scale leaves, while the spur shoots bear the only foli-
age leaves; in Sciadopiiys and Pliyllocladus the same condi-
tion obtains, except that in the former the spur shoot is appar-
ently replaced by the peculiar double needle leaf (Fig. 41),
and in the latter it is transformed into the phylloclad.

An interesting fact in connection with these shoots is that
in the seedlings a very different behavior of the shoots may be
observed. These " juvenile " and usually very transient forms
may be " fixed " by artificial culture, so that the plant assumes

51

FIG. 37.— Sequoia gigantea. This tree is ten meters in diameter.
From a photograph in possession of the United States Geological Survey.

FIG. 38. — Finns Strobus. — After Forest Tree Photo -reproduction Company of Chicago.

MORPHOLOGY OF SPERMATOPHYTES

FIG. 39. — Larix sp., after shedding its leaves. — After COULTER.

a somewhat permanent appearance very different from its nor-
mal adult form. Goebel 46 has recently given an interesting
resume of this subject. For example, in Pinus the long shoots

COXIFE RALES

55

of seedlings bear foliage leaves, which may disappear from the
long shoots in the second year, as in P. silvestris, or may con-
tinue for many years. These needlelike primary leaves are of
simpler anatomical structure than the subsequent foliage leaves,
especially in the matter of provision for controlling transpira-
tion. In the juvenile forms
of Larix the leaves persist
during the winter. It is
among the Cupresseae, how-
ever, that the greatest amount
of work has been done in " fix-
ing " juvenile forms, such
relatively permanent forms
being known in gardens as
species of Retinospora. While
some of the adult forms of Cu-
presseae, as Juniperus com-
munis, retain the more primi-
tive habit of spreading nee-
dles, others (Juniperus Vir-
giniana, species of Cupressus,
of Callitris, of Cliamaecy-
of Thuja) develop
leaves upon

their adult shoots — that is,
leaves whose upper sides have
become organically connect^
ed with the adjacent stem sur-'
face, so that they appear as
<>rof'n scales. In such cases,
the juvenile form, with its
spreading needle leaves, pre-
sents a striking contrast with

the adult form (Fig. 42, .4-#). FlG. iO.—Araucana sp.-Atter COILTEP,

In Pliyllocladus the phyllo-

clads are in the axils of small scalelike leaves which are at first
green, but later become true scales, and are really the trans-
formed primordia of foliage leaves; while in the juvenile forms the
first leaves are flat green needles, which are gradually replaced by
shorter and shorter ones until the adult form is reached, and the
5

pans,

" concrescent

56

MORPHOLOGY. OF SPERM ATOPHYTES

phylloclads themselves only gradually acquire their peculiar
character (Fig. 42, E, F). The juvenile forms of Sciadopitys,
like those of Pinus, show simple needle leaves upon long shoots,
but later scales occur instead of needles, and in their axils the
peculiar double needle leaves are developed.

Goebel suggests that the juvenile forms probably represent
the more primitive form. These facts in connection with the

FIG. 41. — Sciadopitys verticillata : at the left an ovulate strobilus, at the right a branch
with staminate strobili, one third natural size (from Gardener's Chronicle) ; a, ven-
tral view, and ft, dorsal view of staminate sporophyll ; c, ovulate sporophyll ; d, cross
section of the double leaf: a, 6, c, after Sieb. and Zucc. — Whole figure from ENG-
LER and PRANTL'S Nat. Pflanzenfam.

group furnish an admirable illustration of what the zoologists
have styled the " recapitulation theory," which has been tersely
defined as meaning that ontogeny repeats phylogeny. It is cer-
tainly true that if the adult bodies of many of the forms not
mentioned above be associated with these juvenile forms, a
very consistent shoot body is discovered for Conifers. It would
follow that the replacement of foliage leaves by scales on the
long shoots of Pinus, Sciadopitys, and Phyllodadus ; the pecul-

FIG. 42. — Juvenile forms: A-D, Thuja occidentalism A and B, seedlings; G and
shoots from adventitious buds ; E and F, seedlings of Phyllocladus rhomboidalis.

58 MORPHOLOGY OF SPERMATOPHYTES

iar double leaf of Sciadopitys and phylloclad of Phyllodadus',
the deciduous habit of La-rix; the concrescent leaves of many
Cupresseae, are all derived characters, from forms with spread-
ing and persistent needle leaves on all the shoots.

The histological features of the adult stem are quite uni-
form, and in some particulars peculiar to the group. As was
stated under the Cycads, however, the apex of the stem
is variable in character. According to De Bary 27 (page 14),
the dermatogen, periblem, and plerome either merge in a com-
mon group of initials, as in the Abieteae, or these regions are
distinct at the apex, each with its own initials, as in certain
species of Araucaria, Dammara, and Cunninghamia. One of
the most striking features of the stem, in contrast with that of
Cycads, is the persistence of the primary cambium cylinder,
resulting in annual increase of diameter through the accumula-
tion of secondary xylem, a character which constitutes the chief
resemblance of Conifers to Dicotyledons. The primary bundles
contain true tracheary tissue, but writh the appearance of the
secondary xylem the resemblance to Dicotyledons vanishes, as it
consists exclusively of radially arranged tracheids with bordered
pits, and is penetrated by narrow and vertically short pith rays.
In Pinus, Picea, Larix, Pseudotsuga, and probably other gen-
era, the pith rays are of two general types, either very narrow,
often but one cell broad, or lenticular in tangential section,
with a resin duct in the middle. In vertical extent the number
of cells is quite variable, probably three to sixteen representing
the average range of numbers. De Bary 27 (page 490), reports
one to twrelve as the number of cells in the vertical plane of the
pith rays of some Abieteae ; while in Cedrus and in certain fossil
forms the number may rise as high as fifty. The bordered pits
are usually developed only upon the tracheid walls toward the
pith rays, and are discovered, therefore, by radial longitudinal
sections (Fig. 43). The cells of the pith rays also have large
closed pits upon the walls adjacent to th,e tracheids. All of the
bundles are common, a single strand connecting with each leaf,
none of them being cauline as in Cycads. Companion cells are
also lacking, as in Pteridophytes. The open collateral bundles
have retained no trace of a concentric origin, as in the Cycads,
unless the peculiar bundles found by "Worsdell 48 in the cotyle-
dons and fleshy seed integument of C eplialotaxus may be re-

COXIFERALES

59

garded as giving evidence of such a derivation. In other par-
ticulars, also, Ceplialotaxus seems to be the most cycadlike of
the Conifers.

Resin ducts seem to be present in all Conifers except Taxus,
always occurring in the leaves and in the cortex of the stem. In

FIG. 43. — A cube from the secondary wood of Pinus Laricio, reconstructed from three
camera drawings : in cross section at the left a few rows of small thick-walled cells
of the autumn wood are shown, and at the right the larger thinner-walled cells of
the spring wood ; tr, transverse section ; Ir, longitudinal radial section ; It, longi-
tudinal tangential section ; m, medullary ray ; J, bordered pit ; s, simple pit ; x 400.

different genera they occur also in all other regions ; for exam-
ple, in the primary and secondary xylem, and certain pith rays
of Pinus; in the primary and secondary phloem of both stem and
root in Araucaria, etc. They are absent from the root in many
genera, and when present never occur in the cortex. These

60

MORPHOLOGY OF SPERMATOPHYTES

ducts are schizogenous in origin, a group of glandular cells
being formed by repeated division, which separate by the split-
ting of common walls, and organize a passageway which they
line, and into which they pour their characteristic secretion.

The chief histological features of the Conifer stem may be
summarized as follows: (1) growth in diameter by means of
the primary cambium cylinder, (2) secondary xylem radially
arranged and composed entirely of tracheids with bordered pits,
(3) presence of resin ducts, and (4) absence of companion cells.

THE LEAF

The occurrence of foliage leaves and scales is far from uni-
form, as was indicated in the discussion of the stem, and may be
outlined in a general way as follows: (1) foliage leaves only,
as in most Cupresseae and in Araucarieae ; (2) scale leaves only,
as in Phyllodadus (the spur shoots being transformed into
phylloclads) ; (3) both foliage leaves and scales, either on the

FIG. 44.— Taxus laccata : .7, branch with ripe seeds; #, ovule projecting beyond scales
of the small fertile branch ; ,?, longitudinal section of #; ^, young seed not yet in-
closed by aril ; 5, longitudinal section of ripe seed ; <?, Thuja orientalis ; 7-0, Junip-
erus communis. — After KERNER.

same shoot, as in Abies and its allies, or on different shoots, as in
Pinus. It was noted in connection with the stem that spread-
ing needle leaves on all the shoots is probably the primitive
Conifer condition, and that the scales and other leaf forms have
been derived from them.

The foliage leaves are quite variable in form, from small

COXIFERALES

61

discoid ones closely imbricated and appressed upon the axis
(" concrescent "), as in many of the Cupresseae, to the charac-
teristic free needles of the Abieteae, and the broad blades of
certain Podocarpeae (Figs. 44-46, 53). In general the phyllo-
taxy is spiral, but in the Cupresseae it is cyclic. The foliage

FIG. 45. — Phyllocladus spp. : A, branch with staminate strobili ; a, staminate strobilus ;
b and c, sporophylls from a ; d, longitudinal section of ovulate strobilus ; e, mature
ovulate strobilus ; /, longitudinal section of e ; B and <?, ovulate branches. A and
a-d, P. trichomanoides, after HOOKER ; e and /, P. rhomboidalis, after RICHARD ; B
and g, P. glauca. — Entire plate from ENGLER and PRAXTL'S Xat. Pflanzenfam.

leaves are very persistent, enduring from one to ten years, the
basal growth permitting them to increase in size at base with
the increase in the diameter of the axis. So far as recorded,
the only Conifers with deciduous leaves are Larix, Taxodium
(list i chum, and Glyptostrobus, a habit which seems to be a de-
rived one, as in the juvenile forms of Larix the leaves persist
through the winter.

The histology of the foliage leaf reveals a very xerophytic
structure, and argues for adaptation to extreme conditions. The
epidermis consists of elongated, fiberlike cells, with strongly
cutinized walls, the guard cells being deeply sunken. The
rigidity is chiefly due to the hypodermal layers or masses of
elongated sclerenchymatous cells. In case the leaf is flat the
mesophyll differentiates into the palisade and spongy regions,
but if it is acicular the mesophyll is uniform throughout; in
any case there are curious platelike " infoldings " of the wall.
In the acicular leaves there is a single central bundle region,

62 MORPHOLOGY OF SPERM ATOPHYTES

invested by a very distinct bundle sheath, in which two vascular
bundles of the ordinary collateral type are developed. The re-

FIG. 46. — Branch from Podocarpus Purdieana. — From photograph of herbarium speci-
men, two thirds natural size.

maining tissue within the sheath is organized as " transfusion
tissue," consisting of two kinds of parenchymatous cells, (1)
those without protoplasm and pitted, and (2) those with proto-

CONIFERALES 63

plasm and not pitted. The former cells are said to represent
an extension of the tracheid system, passing water from the
xylem to the mesophyll; and the latter to mediate between the
mesophyll and the phloem in the transfer of organized material
(Fig. 47).

The single bundle which enters the leaf divides into two
strands in the narrow leaves, and the strands run parallel and
near together with a common bundle sheath. In broader leaves

*

FIG. 47. — Pinus Laricio, cross section of a needle, x 70.

the bundle divides into several more divergent strands. In no
case do the veins come near the surface, but run through the
middle tissue.

THE ROOT

In strong contrast with the Pteridophytes, the primary root
in Conifers persists as a tap root; while, according to De
Bary 27, following Strasburger,17 the differentiation of the
meristem at the apex is essentially different from that of
Angiosperms, as was indicated in connection with the Cycads.
The description given by De Bary is essentially as follows,
and the accompanying figure (Fig. 48) from the same source
will fully illustrate it: "A plerome cylinder with sharp

64:

MORPHOLOGY OF SPERMATOPHYTES

contour occupies the center. The longitudinal rows of cells
which compose this converge at the rounded apex toward a small

initial group of cells. The ple-
rome is surrounded by a mantle
of periblem consisting of many
(in Thuja occidentalis 12 to 14)
concentric layers arranged with
considerable regularity. Each
one of the inner layers (in
Thuja 8 to 10) of this mantle
has its initial group above the
apex of the plerome." Cell di-
visions perpendicular and paral-
lel to the surface increase the
surface elements of the layer
and the number of layers. " As
the layers are pushed outward
above the apex by their succes-
sive doubling, division ceases in
them, and increase of volume of
the cells takes place ; those which
happen to be outermost at the
apex become gradually loos-
ened, and pushed off as a root
cap. Here, then, it is not possi-
ble to distinguish a layer of calyptrogen or of dermatogen; the
outermost periblem acts as root cap covering the meristematic
apex."

The vascular axis is mostly diarch, a simpler structure than
in Dicotyledons generally, and much simpler tnan in Monocoty-
ledons.

r P

FIG. 48. — Juniperus oxycedrus, longi-
tudinal section of lateral root : p-p,
plerome, surrounded by about sixteen
layers of periblem, the outermost of
which represent the root cap ; *, ini-
tial region for periblem and plerome.
—After DE BAKY.

II. THE SPOKE-PRODUCING MEMBERS

THE MICROSPORAISTGIUM

The strobili of Conifers are always monosporangiate, either
monoecious, as in Abieteae and Thuja, or dioecious, as in Arau-
caria, Juniperus, and Taxus. The microsporangiate strobili
are usually much more numerous than the megasporangiate
ones, and are never terminal on the primary axis, or even upon

COXIFERALES

65

the larger branches (Figs. 49, 50, 53, 57, 58). Sometimes they
appear terminal on small leafy shoots of the last order, and
sometimes they occur in the axils of the leaves of stronger
shoots. In Pinus they replace spur shoots in the axils of scales,
being usually numerous and forming a cluster beyond which
the parent axis continues its growth.

The sporophylls follow the leaf arrangement ; for example,
they are spiral in Abieteae, and cyclic in Cupresseae. They are
exceedingly variable in form, nearly every genus having a char-
acteristic microsporophyll (Fig. 51). In almost every case there
is an evident differentiation into a stalklike base and an expanded
terminal portion, the latter bearing the sporangia. Peltate
forms occur, as in Taxus, recalling the sporophylls of some of
the Cycads and of Equisetuvn, and from beneath the peltate
expansion the numerous free sporangia are pendent by a narrow

''IG. 49. — Pinus Laricio. Longitudinal
section of a staminate strobilus col-
lected in November, x 14.

FIG. 50. — Juniperus comimtnis, longitudi-
nal section of a staminate strobilus,
x 24.

base. Xumerous modifications of the peltate type occur, as in
.raucaria, Cupressus, Cryptomeria, etc., as well as transitions
to the type exhibited by Pinus and its allies, in which the ex-
panded portion continues the plane of the stalk, and bears upon
its lower surface two pouchlike more or less imbedded sporangia.

66

MORPHOLOGY OF SPERMATOPHYTES

A microsporophyll of Pinus bears a certain superficial resem-
blance to the ordinary stamen of Angiosperms, but the " pollen
sac " of the latter is not the morphological equivalent of the
microsporangium of the former. In general, the microsporo-
phylls of Conifers bear less numerous sporangia than do those of
Cycads, and a much more variable number than do those of
Angiosperms.

The sporangia are protected in various ways. Sometimes
they are more or less sunken in the tissues of the sporophyll
(Pinus, Abies, 'etc.) ; sometimes they are covered by an out-
growth from the under surface of the bladelike expansion (Ou-
pressus, Thuja, Juniperus, etc.) ; and in many cases they are
freely suspended by a narrow base, but protected by the close
abutting of the peltate portions.

The sporangium develops in the usual eusporangiate man-
ner, organizing several wall layers and developing numerous
mother cells (Fig. 68). We find that in Pinus Laricio the num-
ber of wall layers is usually five, the innermost organizing the

D

J

K

FIG. 51. — Microsporophylls of various Conifers : A, Taxus baccata ; B, Agathis Dam-
mara ; C, Araucaria Brasiliana, ; J9, Cupressus sempermrens ; E, Cunninghamia
Sinensis ; F, Torreya taxi/olio, ; G, Cedrus Deodara ; H, Sciadopitys verticillata ;
/, Dacrydium cupressinum ; «/, Podocarpus Totara ; K, Pinus Laricio. — A-J se-
lected from ENGLER and PBANTL'S Nat. Pfianzenfam.

peripheral section of the tapetal layer, and the next outer one
being composed of very flat tabular cells, which have probably
been flattened in the process of supplying nutrition to the adja-
cent tapetal cells. Chamberlain 38 has found that the microspo-
rangia of Pinus Laricio, Cupressus Lawsoniana, and Taxus bac-

CONIFERALES

67

cata Canadensis are in the mother-cell stage in October, but how
much earlier they reach this stage he did not discover (Fig. 52).
In this condition the sporangia pass the winter, and during the

FIG. 52. — Winter condition of microsporangia : A, cross section of sporangium of Pinus
Laricio, collected October 1st ; E, same, collected January 3d ; <7, same, collected
April 4th ; Z>, sporangium of Taxus baccata Canadensis, collected October 1st. —
After CHAMBERLAIN.

following spring (about May 1st in the case of P. Laricio) the
mother cells pass through the reduction division.

THE MEGASPORA^GIUM

The structures associated with the megasporangium in Coni-
fers have long been under discussion, and the literature of the
subject is very extensive. There is such great diversity among
the various groups that it is difficult to discover any interpreta-

Fia. 53.— Pinus Laricio, branch from the top of a thrifty tree. June 1st ; small strobili
at extreme top are in stage shown in Fig. 63 ; in the strobili next below, which
are a year older, the archcgonium initials are distinguishable (Fig. 64) ; the strobili
next below, another year older, have shed their seeds ; to the right is a cluster of
staminate strobili, just ready to shed their pollen.

COXIFERALES

69

ii

tion of the structures which can find general application, and
yet the groups are so evidently related that any explanation of
the ovuliferous structures in one group must apply to all. A
brief statement of the principal differences is as follows :

In the Abieteae there is a small bract, a very much larger
and woody ovuliferous scale coalescent with the bract at the very
base, and two inverted basal ovules. In the Araucarieae there
is a prominent bract, a ligulelike ovuliferous scale (present in
Araucarm, wanting in Agathis), and a single imbedded ovule
some distance above the base. In the Taxodieae and Cupresseae
there is a single bract or scale structure with two distinct apices
at the free and enlarged end and bearing one (Juniperus) to
many (Cupressus) ovules; in Juniperus this structure becomes
fleshy in the or-
ganization of the
"berry." In the
Taxeae there is a
single ovule at the
end (?) of a short
secondary leafy
axis, being at-
tached to no scale
or bract, but be-
coming inclosed in
seed by a colored
and fleshy " aril."
In the Podocarpeae (chiefly Podocarpus) there is an " anatro-
pous " ovule associated with a bract but carried up far above
it on a long stalk, and two integuments (the outer becoming
fleshy and colored) ; in some species bracts and axis become fleshy
and fuse together " into a colored, succulent whole " (Figs. 44,
45, 53-58).

The chief contention has been in reference to the nature of
the ovuliferous (or seminiferous) scale, and in reference to the
integument. A very valuable historical study of the subject
has been published recently by Worsdell,47 who has brought
together in compact form the principal views which have been
held, and to whom we are indebted for much in the following
brief account.

Before 1827, in which year Eobert Brown * announced gym-

FIG. 54, — Finns Laricio, diagram of ovule and associated
structures ; J, bract ; s, ovuliferous scale ; f, vascular
bundles ; m, micropyle ; *', integument ; />, pollen grains ;
c, cavity in apex of nucellus ; », nucellus ; g, embryo sac.

70

MORPHOLOGY OF SPERMATOPHYTES

nospermy, the ovuliferous structures were interpreted in terms
of Angiosperms, the ovule being regarded as a pistil, and the
related parts being variously interpreted. Brown's conclusion
as to a naked ovule was derived from a comparison of the so-

FIG. 55.— Ovulate structures of various Conifers: 1, Agathis aitstralis, ovuliferous scale
from inner side (Jf, winged seed) ; #, longitudinal section of 1 ; <?, Araucaria excelsa,
longitudinal section of scale, etc., also showing the outgrowth (i) above the seed ;
4, Cunninghamia Sinensis, ovuliferous scale, showing three ovules ( M), and an
outgrowth (i) ; 5, Mierocachrys tetragona, longitudinal section of ovuliferous scale,
also showing the arillus (a) and the outgrowth (i) ; £, Oryptomeria Japonica, longi-
tudinal section of part of the strobilus ; 7 and #, Cupressus Lawsoniana, showing a
young cone (7) and a later stage (8) ; £, Podocarpus macrophylla, longitudinal sec-
tion, showing ovulate structures and aril (ar). — From ENGLER and PRANTL'S Nat.
Pfianzenfani.

called " ovule ?? (nucellus) of Cycads and Conifers with the
ovule of Angiosperms. His corollary was that the ovuliferous
scale represents an open carpel, but his statement that this so-
called carpel is a leaf in the axil of a bract called forth strong
dissent.

In 1839 Schleiden 2 called attention to the fact that Brown's
" folium in axilla folii " is a morphological impossibility, and
that the ovuliferous scale is a flattened axis in the form of a
placenta, a view concurred in later by Baillon, Dickson, Stras-
burger, and Masters, but without regarding the axis as a pla-
centa, the axial nature of placentas in general being one of
Schleiden's peculiar views.

COXIFERALES

71

In 1845 Von Mohl 3 suggested that the ovuliferous scale is
a leaf of an axillary shoot, but later (1871) he modified his view
somewhat.

In 1853 A. Braun 4 first advanced the theory that the
ovuliferous scale represents the first two leaves of an axillary
shoot, which are fused by their margins, a view held later by
Caspary, Parlatore, Oersted, Yon Mohl, Stenzel, Engelmann,
Willkomm, and Celakovsky. Braun's illustrative material con-
sisted of a monstrous cone of Larix, in which the ovuliferous
scale was replaced by a short branch bearing two leaves trans-
versely placed, the bract developing as a foliage leaf.

It may be of interest to note that in 1860 Baillon 6 an-
nounced his opposition to the theory of gymnospermy, a posi-

FIG. 56.— Ovu^ate structures of various Conifers : 1, Abies pectinata, ovulate strobilus ;
2. ^dorsal view of bract and ovuliferous scale ; #, ventraT view of same ; 4i longi-
tudinal section ^f same ; 5, a winged seed ; £, longitudinal section of seed ; 7, Pinus
silvestris, ventral view of ovuliferous scale ; 8, Larix Europaea, ovuliferous scale,
and bract with bristle ; 5, longitudinal section of same. — After KERXER.

tion which he maintained persistently, basing it upon the first
really careful researches in the organogeny of the structures
under discussion. He sustained Schleiden's view that the ovu-

72 MORPHOLOGY OF SPERMATOPIIYTES

liferous scale is an axis, but regarded it as an axillary shoot
rather than a placenta.

In the same year (1860) Dickson8 recorded some cones of
the Norway spruce (Picea excelsa) in which the lower bracts
were replaced by stamens, while the upper bracts bore axillary

FIG. 57. — Cephalotaxus spp. : staminate branch and details of structure (a-g) of C. For-
tunei ; ovulate branch of C. pedunculata. — From ENGLER and PKANTL'S Nat. Pflan-
zenfum.

scales as in the normal ovuliferous cone. His position, there-
fore, was that stamens are to be homologized with the bracts
of the ovuliferous cone, and that the ovuliferous scales repre-
sent axes of the next higher order, and are " simple flattened
shoots," as Schleiden had urged. It may be remarked that

COXIFERALES

73

a similar monstrosity has been reported by Shaw 34 in con-
nection with the cones of Sequoia. For some, such illus-
trations settle the sporophyll character of the bract, which
to them therefore becomes the carpel in the ovuliferous cone
(Figs. 59, 60). . p.

In 1860 also Caspary 7 confirmed Braun' s conclusion, citing
abnormal specimens in which branches occurred in the axils of
the brac.ts, and ' bearing
the two- halves of' the
ovuliferous scaje as lat-
eral appendages.

In 1864 Parlatore n
recorded a monstrous
cone of Finns Pinaster
(P. Lemoniana) in which
an ordinary spur shoot
with its two needle leaves
sprang from the axil of
every alternate bract, re-
placing an ovuliferous
scale. His conclusions'
naturally accorded with
those of Braun.

In the same year
(1864) Oersted 10 * re-
corded some remarkable
cones, in which the low-
est bracts had the form
of foliage leaves; in the
axils of the next higher
ones were several scales

as on a suppressed axis, the two outermost being largest and
opposite ; higher up the bracts became gradually smaller and the
axillary scales less numerous, but the two outermost scales grad-
ually increased in size and became connate by their posterior
margins, while rudimentary ovules appeared at their base; in
the uppermost part of the cone the bract was reduced to its ordi-
nary size, and the ovuliferous scales had fused into a single large
broad structure dentate or bifid at the apex. His conclusion
in general was that of Braun, but in his detailed application he

FIG. 58. — Podocarpus spp. : staminate and ovulate
branches from P. Totara ; a, microsporophyll ;
i, ovulate structures ; c, longitudinal section of
same ; d, ovulate structures from P. macrophyl-
la ; «, longitudinal section' of same. — Branches
and a-c after HOOKER ; whole plate from ENG-
LER and PRANTL'S Nat. Pflanzenfam.

74 MORPHOLOGY OF SPERMATOPHYTES

regarded the ovuliferous scale in Abieteae as an open carpel,
which is lacking in Cupresseae.

One of the simplest explanations of these problematical
structures was that proposed by Sachs 13 in 1868, and afterward
elaborated more fully by Eichler.18 They contended that the

FIG. 59. — Abnormal strobili of Abies sp. : the staminate cone at lower right-hand corner
is almost normal ; all the others have microsporangia both above and. below, with
ovuliferous scales between.

bract is a carpel and the ovuliferous scale a ligular outgrowth
from it, calling attention to a similar condition of things in the
leaves of Isoetes and Selaginella. Such a ligular placenta does
not appear in certain groups, as Cupresseae, Taxodieae, and Podo-
carpeae, and in these cases the bract is evidently an open carpel.
The simplicity of this explanation has commended it to writers
of text-books, and hence it is perhaps the view which is most
current.

In 1869 Van Tieghem 14 presented his conclusions based
upon anatomical structure — a new point of view. He claimed
that the ovuliferous scale is the first and only leaf of an axillary
branch, although he suggested the possibility of two fused leaves,
and bases his statement on the course and orientation of the
bundles. He states that the bundles of bract and ovuliferous
scale leave the main axis each in its own sheath, and thus repre-
sent independent systems of bundles ; that the uppermost
divides and forms an arc with inverted orientation, the arrange-

CONIFERALES

75

merit of bundles in the arc showing that the axillary structure
is a leaf and not a branch; and that the reversed orientation
shows that the leaf belongs upon the suppressed branch opposite
the bract. This is not Robert Brown's view of an axillary car-
pel, but rather Alexander Braun's view that the ovuliferous
scale is the leaf of an axillary branch.

In 1871 Von Mohl 16 further strengthened Braun's position
by the publication of his studies of the peculiar " double leaf "
of Sciadopitys. He showed that this leaf represents the first
two leaves of an axillary shoot, which stand transversely, and
which become coalescent by their posterior (toward the axis)
edges, the vascular bundles thus showing reversed orientation.
As this normal behavior of the leaves of Sciadopitys exactly
parallels what was claimed by Braun for the ovuliferous scale,
the results of Von Mohl are almost in the nature of a demon-
stration.

In 1872 Strasburger 17 announced his adherence to the view
that the ovuliferous scale is a flattened axis, at the same time
combating the idea of gymnospermy. While later 24 his views

A B c D & F

FIG. 60.— Microsporangia from the strobili shown in Fig. 59: A, normal sporophyll,
side view ; B, slightly modified sporophyll from summit of a bisporangiate stro-
bilus ; C and Z>, side and front views of sporophylls from summit of a bisporangiate
strobilus ; E and F, side and front views of sporophylls from the base of a bispor-
angiate strobilus.

in reference to gymnospermy became modified, he held to the
axial nature of the scale.

In 1876 Stenzel 20 described striking abnormalities in Picea
excelsa. In one cone leafy axes occurred in the axils of the
bracts, the first two leaves resembling ovuliferous scales more
than ordinary foliage leaves in texture. In other cones the
abnormalities recorded bv Dickson 8 were observed. In still

70 MORPHOLOGY OF SPERMATOPIIYTES

other cones he found the two parts of the ovuliferous scale in
all stages of coalescence. All of his material confirmed Braun's
view that the ovuliferous scale is made up of the first two leaves
of an axillary shoot, which stand transversely, and are connate
by their axial edges. It may be of interest to note in this con-
nection that such a view would refer the megasporangia to the
niorpholpgically under surface of the leaf, and in so doing would
make them conform with the microsporangia in this respect.

In the same year (1876) Engelmann,19 in reviewing Sten-
zel's paper, reported similar monstrosities in cones of Picea
Engelmanni and Tsuga Canadensis.

In 1879 Celakovsky 23> 25> 29> 35 began publishing upon the
subject, and has constructed a theory which seems to unify more
nearly than any other the puzzling diversities of structure. In
the first place, he regards the ovuliferous scale as the repre-
sentative of an axillary shoot, confirming in general Braun's
view. He states that in this case " Nature takes a short cut,"
the ovuliferous scale being formed directly, rather than an axis
with two distinct lateral leaves. Moreover, he sees in the ovu-
liferous scale, which has assumed a permanently vegetative char-
acter, the modified and blended outer integuments of the two
ovules. Therefore, there is present no true carpel in Conifers,
this structure being reduced, so to speak, to a single ovule. The
microsporophyll corresponds to the bract of the megasporangiate
cone, which subtends the axis which bears the hypothetical
megasporophyll (represented only by the ovule), the latter
being on a shoot of higher order. As a consequence, the axillary
ovuliferous structure corresponds to the axillary spur shoot of
Ginkgo and the main axis of Cycads. In applying this doc-
trine of an outer integument in all Conifers, Celakovsky reaches
the following conclusions: In the Taxaceae, certainly the older
of the twro great groups of Conifers, the two normal integu-
ments are evident, being distinct in Taxus, Dacrydium, and
Microcachrys (the outer being the " aril "), and represented by
the fleshy and bony layers of the seed in Cephalotaxus and
Podocarpus, as in Cycads and Ginkgo. In Araucarieae, Abie-
teae, and Cupresseae, the outer integument is the ovuliferous
scale (or half of it), being developed after the sporangium and
showing the inverted orientation, facts true of all outer integu-
ments. In this outer integument he sees the ligule of Isoetes

CONIFERALES 77

and SelagineUa, and in the inner integument the indusium of
Perns and velum of Isoetes.

It would seem as though all possible explanations had been
suggested. Upon sifting the testimony certain things seem to
be fairly clear, and one is that the scale and its ovules in Abie-
teae represent a highly modified axillary shoot, corresponding to
the characteristic spur shoot of the group. In Taxus, repre-
senting the more primitive forms, this ovuliferous spur shoot
develops normally, bearing a few leaves, and organizing its
ovule in a terminal position, but really in the axil of the upper-
most leaf, a fact further emphasized by the lateral and axillary
ovule of the allied Torreya. In these cases the ovule is dis-
tinctly cauline in origin, and holds the same relation to the sub-
tending bract as does the sporangium of SelagineUa. Perhaps
too much stress has been laid upon the relation of the ovule to
the member which produces it. It may be developed from the
upper or under surface of the leaf, or from the axis, in the same
alliance, but always from periblem tissue. One may expect all
of these external expressions of its origin in so ancient and
diversified a group as the Conifers, and the desire for absolute
uniformity in this regard will probably not be realized. The
same spur shoot appears in a somewhat elongated and naked
form in Cephalotaxus, while in Podocarpus it is more or less
leafy, the uppermost leaf or two bearing an axillary and stalked
ovule. Among the Pinaceae the ovuliferous spur shoot reaches
its extreme modification, being represented only by its first two
leaves, which have fused to form the ovuliferous scale, and this
scale may be more or less coalescent with the subtending bract.
Whether those leaves of the spur shoot which are intimately asso-
ciated with ovules, either bearing or subtending them, are to be
regarded as carpels or as modified outer integuments is a ques-
tion we are in no position to answer. It seems simpler to regard
them as carpels — that is, leaves modified to bear or subtend
ovules, and such a view will be adopted in this book, though not
strenuously insisted upon.

Our knowledge of the development of the mega sporangium
and of the megaspore in Conifers does not permit generalization,
and it is impossible to give an account which will include the
whole group. One remarkable feature in the forms studied,
which may belong to the group in general, is the extreme slow-

78

MORPHOLOGY OF SPERMATOPHYTES

ness with which these structures develop. We have obtained
exact information concerning this feature in Pinus Laricio. In
the spring, young ovules with distinct integument and nucellus
are found, but with no apparent differentiation of sporogenous
tissue. It should be remarked that at this season Strasburger 24
found the mother cell differentiated in Larix, and suggested
that the ovule had passed the previous winter in this condi-
tion. In May the mother cell in Pinus becomes very apparent
through great increase in size. It is this stage which is to be

FIG. 01. — Larix Europcea : A, longitudinal section of a young ovule, showing the mega-
spore mother cell and tapetal cells, also at the left the beginning of the integument
(March 1st) ; B, first division of the mother cell ; (7, the three cells derived from
the mother cell ; /), the beginning of the free nuclear division within the func-
tional megaspore ; x 150. — After STRASBURGEK.

observed in the youngest evident cones, such as appear at the
very top of Fig. 53. In the following October the endosperm
has begun to develop, and is found as a parietal cytoplasmic
layer with few to numerous imbedded nuclei and a central
vacuole (Fig. 62), and in this condition the second winter
is passed. In the following spring the endosperm begins
to develop rapidly, and in June the archegonia are ready for

COXIFERALES

79

fertilization, which occurs about the 1st of July, at least tweiity-

oiie months after the first organization of the ovule. It should

also be remembered that after fertilization the seed is not shed

until the following year, fully three years

after the first appearance of the ovule.

Practically the same history was reported

by Strasburger 24 in the case of Larix, and

it probably represents a common condition

among Conifers, a condition which is hard

to detect without the greatest care.

The portion of this long history which
concerns us at present ends with the dis-
tinct organization of the mother cell, and
it is this very period which seems to have
received the least attention. There is cur-
rent the general statement that the arche-
sporium is differentiated very early as one
or more hypoderinal cells. While the
probabilities are largely in favor of the
truth of this statement, we have discovered
no complete series of figures to substan-
tiate it except those of Larix given by
Strasburger.24 In the same connection he
gives certian figures of Taxus which indi-
cate the same fact, and also remarks that
Thuja, Finns silrestris, and Pinus Pu-
milio are essentially the same as Larix in this regard. It should
be remembered, therefore, that the only close series is from
Larix , and that the series is the same in other forms is more
or less a matter of inference from fragmentary observations
(Fig. 61).

A periclinal wall divides each archesporial cell into an outer
primary wall cell and an inner primary sporoge'hous cell. The
wall cells divide repeatedly, and as these divisions are accom-
panied by division of the overlying epidermal cells .an extensive
mass of sterile nucellar tissue is developed between the micro-
pyle and the sporogenous cells. Moreover, the divisions are so
regular that definite rows of sterile cells extend from the sporog-
eiious tissue to the tip of the nucellus. So far as recorded, the
nucellus in Conifers does not develop the persistent beak char-

FIG. 62. — Pinus Laricio,
condition of megaspo-
raugium on May 1st : «',
integument ; p, pollen
tube ; », nucellus ; g,
beginning of gameto-
phyte, the embryo sac
being surrounded by a
definite region of break-
ing down tissue.

80

MORPHOLOGY OF SPfiRMATOPHYTES

acteristic of Cycads and Ginkgo, and the extinct groups ; nor is
there any distinctly organized pollen chamber, although we
have observed in Pinus that the nucellus breaks down at the
apex, so that the pollen grains lie in a cuplike depression.

Although the primary sporogenous cells are cut off by the
hypodermal layer they are not distinguishable from the neigh-
boring cells until they have become deeply placed in the chalazal

FIG. 63.— Pinus Laricio, longitudinal section of megasporangia June 1st, with mega-
spore mother cell in the center, surrounded by a region of more or less modified cells ;
the nucleus of the mother cell is in the prophase of the reduction division ; x 500.

region of the nucellus, and even then there is no differentiation
to attract attention until a mother cell begins to enlarge. This
tardy differentiation of the sporogenous cells in appearance was
remarked by Strasburger in the case of Taxus, and we have
found it to be strikingly true in Pinus. Apparently the pri-
mary sporogenous cells do not divide, but one or more of them
pass over directly into mother cells. Several mother cells have
been observed to begin to enlarge in Taxus and in Sequoia, but

COXIFERALES 81

in Larix and Pinus never more than a single one has been ob-
served to become distinct from the neighboring cells.

In our figure, of a mother cell of Pinus Laricio imbedded in
nucellar tissue (Fig. 63), it is apparent that it is surrounded by
a rather definite zorie of cells, two to four layers in depth, which
give evidence of breaking down. After endosperm formation
is somewhat advanced, this investing zone becomes differentiated
into two distinct regions, an outer layer of tabular, almost empty
cells, and an inner region of polygonal cells with densely stain-
ing contents. It is this appearance which has sometimes led to
the impression that a definite tapetum surrounds a sporogenous
mass. The most usual mistake, however, is to regard the dis-
tinct mass of polygonal cells as endosperm cells, overlooking the
comparatively inconspicuous embryo sac, a mistake made by
Hofmeister 9 and corrected by Strasburger.24

A£ the mother cell enlarges, its nucellus begins to show signs
of the organization which precedes the reduction division and
would result in the first appearance of the gametophyte. It
may be of interest to note that after the mother cell has begun
to enlarge it becomes full of starch, but by the time it has
reached the spirem stage for division the starch has disappeared.
How far the history outlined above is uniform throughout Co-
nifers must be left to subsequent investigation to determine.

In Sequoia Shaw 34 has observed an interesting modifica-
tion. When the growth of sterile tissue at the apex of the
nucellus has covered the sporogenous cells with five or six layers,
the sterile tissue of the nucellus beneath the sporogenous region
begins rapid division and elongation, again placing the spo-
rogenous tissue relatively high in the nucellus.

III. THE GAMETOPHYTES

THE FEMALE GAMETOPHYTE

The female gametophyte begins with the reduction division
of the megaspore mother cell. In the cases investigated, the
mother cells seem to be the primary sporogenous cells, which
form a more or less extensive group, and which, although origi-
nating from hypodermal cells, have become deeply placed beneath
the extensively developed sterile wall tissue. The deeply placed
mother cells (one in Larix and Pinus, usually more in Taxus

82 MORPHOLOGY OF SPERMATOPHYTES

and Sequoia) , after considerable increase in size, presumably
pass through the reduction division, although this has never been
definitely observed. The reduction number of chromosomes
was observed by Dixon 33 in the endosperm cells of Pinus sil-
vestris', while Blackman 37 and Chamberlain39 definitely
counted twelve chromosomes in the endosperm of Pinus silves-
tris and Pinus Laricio respectively. In the case of Larix Stras-
burger 24 has demonstrated the existence of a row of three cells
derived from the mother cell (Fig. 61), and in Taxus the same
number and sometimes more, in both cases the lowest cell becom-
ing the fertile megaspore. The same investigator makes the
general statement that Thuja, Pinus silvestris, and Pinus Pu-
milio are essentially similar. We have recently discovered this
stage in Pinus Laricio (Fig. 106), too late to include the figure at
this place. A row of four potential megaspores is very evident,
the lowest one becoming functional, the other three appearing
for a time as a densely staining cap. In the preparation figured
the original wall of the mother cell is sharply defined, but the
walls between the disintegrating cells are very indistinct.

In case several mother cells are functional, as in Taxus and
Sequoia, several megaspores may begin to enlarge simultane-
ously, but one soon dominates, and enlarging at the expense not
only of the other megaspores, but also of the adjacent sterile
tissue, becomes the single very large fertile megaspore of the
sporangium. Sometimes the sterile megaspores are somewhat
prominent in the nucellus, and in Sequoia Shaw 34 records them
as clustered about the upper third or fourth of the remarkably
elongated functional megaspore.

The usual account of the germination of the megaspore and
the development of the archegonia has been derived from the
publications of Hofmeister 9> 12 and Strasburger,15- 17> 21> 24- 28
but recently much has been added in the way of detail. Very
early in its history, long before it has attained its fertilization
size, the megaspore begins to germinate, and embryo sac * and
gametophyte continue growth together until the nucellus is
largely replaced. The sequence of events is much the same as

* For convenience we make an arbitrary distinction between megaspore
and embryo sac, the former being the true uninucleate spore, the latter the
persistent and growing wall of the megaspore, which continues to invest the
gametophyte.

CONIFERALES 83

that given for Cycads. The megaspore nucleus divides, and
this is followed by repeated simultaneous nuclear divisions
until a large number of free nuclei have been organized.
According to Jager,40 about 256 free nuclei appear in Taxus
baccata before walls are formed. This represents eight suc-
cessive divisions, and coincides exactly with Hirase's estimate
of the free nuclei in the embryo sac of Ginkgo previous to
the formation of Avails. Very early in this series of divisions,
probably in general when but two or four free nuclei have ap-
peared, the nuclei become imbedded in a parietal cytoplasmic
layer inclosing a large central vacuole (Fig. 62), and in this posi-
tion free nuclear division is completed. Subsequent divisions of
the nuclei are accompanied by the formation of cell walls and a
parietal tissue is organized. From the parietal plate of cells
the cavity of the embryo sac is gradually filled with tissue, the
faces of the cells directed toward the center in this advance lack-
ing walls. The more compact tissue toward the micropyle or-
ganizes archegortia, while the deeper tissue is nutritive.

A deviation from this method of endosperm formation has
been reported by Arnold 42 as occurring in Sequoia. After the
development of the free nuclei, the cytoplasmic layer increases
in thickness, accumulating especially at the lower end of the sac.
Then cell formation occurs in the upper and lower extremities of
the sac, while free nuclear division continues in the central
region.

About the last of May the archegonium initials become ap-
parent. "Within a week the primary neck cell has been cut off,
and the central cell has become much enlarged (Fig. 64, A, B).
As the archegonia develop, the adjacent endosperm tissue con-
tinues to grow, so that a more or less evident depression appears
over each archegonium in case they are scattered (Abieteae), or
over each group of archegonia in case they are clustered (Cupres-
seae and Taxodieae). The number of archegonia developed by
a single gametophyte varies considerably. In Abieteae there are
three to five; in Taxus baccata five to eleven; in Cupresseae five
to fifteen, or even thirty. The neck of the archegonium shows
much variation. In Cycads and Ginkgo the primary neck cell
divides but once, by an anticlinal wall, forming a neck of two
cells, placed side by side as are the guard cells of a stoma. In the
majority of Conifers observed a plate of four cells is derived

84 MORPHOLOGY OF SPERMATOPHYTES

from the primary neck cell, and by periclinal divisions the single
plate becomes two, so that the neck is composed of two tiers of
cells with four cells in each tier. The recorded deviations are
as follows: in Tsuga Canadensis, 31a, 49, and C eplialotaxus
Fortunei the neck is usually two-celled, as in Cycads and Ginkgo',
in some Cupresseae eight cells occur in each tier; while in some
specis of Pinus and Picea more than two tiers of cells occur, with
four or eight cells in each tier.

After the separation of the neck cell from the central cell,
the latter begins a remarkable increase in size and receives a

B ^m^^fw c D

FIG. 64. — Pimis Laricio : A, archegonium initial, May 28th ; B, neck and central cells,
June 2d ; <?, central cell just before cutting off the ventral canal cell, June 18th ;
/>, cutting otfof the ventral canal cell, June 21st ; x 104.

large amount of nutritive material (Fig. 64, (7). About it a
definite jacket of endosperm cells is organized, resembling the
venter of an archegonium. The cells of the jacket become sur-
charged with protoplasmic material, and their nuclei enlarge,
while the wall of the central cell becomes thick and very dis-
tinctly pitted. Through these pits Goroschankin 26 first traced a
continuity of cytoplasm between central cell and jacket cells.
Very early in its history, at the beginning of its enlargement, the
central cell becomes vacuolate, the protoplasmic contents forming
merely a wall layer, the nucleus retaining its apical position from

CONIFERALES 85

the time of the cutting off of the neck cell to the cutting off of
the ventral canal cell, a period of two or three weeks, beginning
between the middle of May and the 1st of June, dependent on
the season. The central cell now begins to receive supplies from
the jacket cells, which practically empty themselves through the
pits. As a consequence, the central cell becomes packed with
food material, the cytoplasm being filled with large and deeply
staining masses, even more prominent in the egg, the best organ-
ized of which have been called " proteid vacuoles." These " vac-
uoles " of the central cell and egg were regarded by Hofmeister
and Goroschankin to be nuclear structures, but Strasburger de-
nied their nuclear nature, and claimed that they are proteid
vacuoles, and such they have been regarded ever since, until the
recent investigations of Arnoldi.43 He observed that the nuclei
of the jacket cells become amoeboid, squeeze through the pits,
and regain their form in the central cell and egg. Xuclei were
also observed to pass from the next outer layer into the cells of
the jacket. In addition to these organized nuclei, nuclear frag-
ments and other materials pass in from the jacket cells, which
may often be seen emptied of their contents.

The cutting off of the ventral canal cell has been known in
a general way since Strasburger 24 established its existence in
Juniper us. The fullest accounts, however, have been given by
Blackmail 37 and by Chamberlain.39 Just before fertilization,
which we find in Pinus Laricio to occur about the 1st of July,
the ventral canal cell is cut off (Fig. 64, D). A broad spindle,
which is almost rectangular in outline before it becomes bipolar,
is organized at the apical extremity of the central cell, and there
appears a thin but broad and sinuous cell plate. This cell plate
cuts out from the main mass a small oval segment, the ventral
canal cell (Fig. 65). After the separation of the ventral canal
cell, its nucleus begins to return to the resting stage, but accord-
ing to Blackman it never reaches it before disorganization sets
in. After this the ventral canal cell appears in various stages
of disorganization, often separating somewhat from the egg, and
has usually disappeared by the time the pollen tube has pene-
trated the neck of the archegonium, although Chamberlain
found occasional traces of it even after the embryo was con-
siderably advanced.

Blackman noted the exact similarity between the nuclei of

86

MORPHOLOGY OF SPERMATOPIIYTES

the ventral canal cell and of the egg. He called attention to
the fact that the two cells differ from each other only in the
quantity of cytoplasm, and concluded that the ventral canal cell

FIG. 65. — Pinus Laricio : A, the ordinary spindle in cutting off the ventral canal cell ;
B, disorganizing ventral canal cell ; C, an unusually large spindle in cutting off
ventral canal cell, the two nuclei being of the same size and showing the same de-
velopmental changes : Z>, a large ventral canal cell ; E, the wall separating a large
ventral canal cell and the egg has broken down (n. »., nucleus of ventral canal cell ;
n. 0., egg nucleus) ; A-D, x 500. — After CHAMBERLAIN.

is evidently an arrested gamete. Chamberlain further strength-
ened this view by discovering a series in which a ventral canal
cell very much larger (" eight to ten times ") than usual was cut
off (Fig. 65, (7, .D). Its nucleus had reached the size of the egg

COXIFERALES

87

nucleus, and had passed through the same peculiar developmental
changes. To all appearances it was ready for fertilization. The
same observer also discovered cases of large ventral canal cells
and nuclei in which the cell plates separating them from the egg
had disorganized, giving the appearance of two similar nuclei
imbedded in the cytoplasm of the egg (Fig. 65, E). It is very
probable that this may explain such cases of equal nuclei as
those recorded by Stras-
burger,22 Coulter 36 (Fig.
66), and others, which
were thought to be the
male and female nuclei
approaching for fusion.
In any event, it seems
clear that the ventral
canal cell represents an
abortive egg, which is oc-
casionally organized as an
egg, and which may rare-
ly function as one.

In a recent paper by
Arnoldi 43 upon Cephalo-
taxus Fort une i a curious
procedure -in reference to
the ventral canal cell is
recorded. The nucleus of
the central cell divides as
usual, and a ventral canal
cell nucleus is formed,
but no regular cell is or-
ganized. Instead of this,
the upper part of the cy-
toplasmic mass, contain-
ing the ventral canal cell
nucleus, is said to become mucilaginous, breaks down the neck
cells (two in number), and is separated from the egg by a pinch-
ing-off process. Although there is no real cell formation in
the morphological sense, physiologically the result is the same as
in other cases, a nucleus and some cytoplasm being separated
from the forming egg.
7

FIG. 66. — Pinus Laricio^ conjugation of two nuclei
described as male (m) and female (/), but
which may prove to be the nuclei of the ven-
tral canal cell and the egg (as in Fig. 65, E) ;
x 500.— After COULTER.

88 MORPHOLOGY OF SPERMATOPHYTES

At the time of its organization, the egg nucleus is no larger
than that of the ventral canal cell, or of the nutritive jacket
cells, but it begins immediately a very rapid and a very great

*wm

A ^sg/;:v;vv^>

FIG. 6Y. — Finns Laricio : Two nuclei showing the collection of the chromatin nucleoli.
into a definite area; x 500.— After CHAMBERLAIN.

enlargement. It moves towards the center of the egg with great
rapidity, and attains a size that seems out of all proportion to
its original bulk and to the time employed. During this en-
largement changes occur in the nucleus which have been differ-
ently interpreted, but which are decidedly at variance with the
cytological phenomena observed in the eggs of Angiosperms.
The chromatin seems to be very scanty and the nucleoli become
numerous. Chamberlain 39 has concluded that the chromatin
takes the form of nucleoli, which finally collect from all parts of
the nucleus to a definite area near the center, where they
take the form of elongated masses that undoubtedly represent
the chromatin of the nucleus (Fig. 67). In the cytoplasm of
the egg the nutritive materials derived from the jacket cells
become prominent, and among them the nuclei (" proteid vac-
uoles ") are conspicuous.

COXIFERALES

89

THE MALE GAMETOPHYTE

The male gametophyte of Conifers has long been known in a
general way, but Belajeff,30 investigating Taxus baccata and
Juniperus communis in 1891, seems to have been the first to
recognize the homologies and functions as at present understood.
In 1892 Strasburger,31 studying a somewhat wider range of
forms, notably Larix and Cupressus, extended BelajefFs con-
clusions to Conifers in general. This was followed in 1893 by
the further studies of Belajeff32 upon Taxus and Juniperus.
In 1894 Dixon 33 published a detailed account of the male
gametophyte of Pinus silvestris, which was followed by us in
1896 with confirmatory studies of Pinus Laricio. In 1896
Shaw 34 published partial results of his studies of Sequoia ;
in 1899 Jager 40 investigated Taxus baccata anew, and with
abundant material ; recently Coker 45 has published some notes

FIG. 68.— Pin t(# Laricio, showing mitosis in pollen mother cells, May 3d ; x 500.

on Taxodium distichum ; while Arnoldi 43 has obtained some
interesting results from a study of Cephalotaxus Fortunei.
These accounts indicate a fairly uniform series of events, a uni-
formity somewhat obscured by the varying terminology of the
writers. We have secured a very complete series of prepara-
tions from Pinus Laricio, and the following account is written
from them (Figs. 68, 69).

90 MORPHOLOGY OF SPERMATOPHYTES

The microsporangium passes the winter in the mother-cell
stage. About the 1st of May the mother cells are found in vari-
ous stages of the reduction division. The divisions do not
appear to be so simultaneous as in the Angiosperms, for in the
same sporangium some mother cells are preparing for the first
division, others contain completed tetrads, and others represent
every stage between (Fig. 68). In the same sporangium, also,
the division may be simultaneous or successive, although the
former seems to be the prevailing method; and they may be
tetrahedral or bilateral. As was stated in connection with the
female gametophyte, the reduction number of chromosomes
in Pinus silvestris and Pinus Laricio has been found to be
twelve. In 1892 Strasburger 31 counted twelve chromosomes in
the endosperm of Pinus silvestris. Recently Blackman 37 has
counted the same number in the oosphere, the jacket cells, the
ordinary endosperm, and the pollen mother cells ; while in the
first division of the oospore, and in the embryo, twenty-four were
counted. Chamberlain 39 obtained the same results from Pinus
Laricio, counting twelve chromosomes in the jacket cells, the
ordinary endosperm, and the pollen mother cells. The wings of
the spores begin to develop while they are within the mother cell.

About May 20th the nucleus of the spore enlarges for its
first division (Fig. 69, D\ a spindle is formed rapidly (Fig.
69, E), and an equal division follows, so far as the mass of
chromatin is concerned. Before the cell plate is organized,
however, the nucleus nearer to the wall of the spore begins to
disorganize, and the other begins to enlarge (Fig. 69, F). In
this way a lenticular and disorganizing cell is cut off against
the wall of the spore (Fig. 69, G). A second division immedi-
ately follows, the spindle being observed about May 25th (Fig.
69, H). This division is a repetition of the first in details, and
the two lenticular cells disorganize rapidly (Fig. 69, I\ become
flattened against the spore wall, and very soon appear merely
as two thin and darkly staining disks (Fig. 69, J).

It seems reasonable to suppose that these two evanescent
cells represent a vestige of the vegetative tissue of the gameto-
phyte, and they may be called properly vegetative cells. It will
be remembered that in the Cycads only one such cell is said to
appear, and that it persists; while in Ginkgo two appear and
the second one persists.

COXIFERALES

91

The third division immediately follows (Fig. 69, J"), divid-
ing the large antheridial cell into the primary spermatogenous

L

FIG. 69. — Finns Laricio, showing a series from the formation of the tetrads to the de-
velopment of the pollen tube : p, vegetative cells ; .<?, stalk cell ; 5, body cell ; t, tube
nucleus : A and B, May 3d ; <7, May 10th ; D-G, May 20th ; H-J, May 25th ; K,
June 15th ; Z, May 1st (nearly a year after stage shown in K} ; x 600.

cell or generative cell, and the tube cell (the so-called " vege-
tative cell") (Fig. 69, K). Until 1891, when Belajeff30 dis-

92

MORPHOLOGY OF SPERMATOPHYTES

covered the true functions, the more prominent nucleus of
the tube cell was regarded as the generative nucleus. By
this time the large tube cell has become very full of starch,
and no further division occurs until
the following spring, a period of
about eleven months. In this con-
dition also pollination occurs, and
the old spore wall with its con-

FIG. 70. — Pinus Zarm'c>, showing flattened
end of pollen tube before discharge into
the egg ; the two male cells are at the
tip and above them, but uot shown,
are the nuclei of the stalk and tube
cells ; x 500. — After CHAMBERLAIN.

FIG. 71. — Pinus Laricio, showing end of
pollen tube just before discharge ; the
male cells (w, m) in this case are be-
hind the nuclei (ri) of the tube and
stalk cells ; s, starch grains ; the pit is
evident at the extreme tip ; x 500. —
After COULTER.

tained gametophyte rests in the cuplike depression at the apex
of the nucellus during all this period.

Dixon 33 says that in Pinus silvestris the pollen tube is sent
out into the nucellus as soon as the grain is deposited, and that
it passes a year or even more in the upper region of the nucellus,
disorganizing the adjacent tissue and producing irregular cav-
ities. This seems to be true of Pinus Laricio also, whose pollen
tubes we have found in the nucellus in June. Coker reports
that in Taxodium distichum, however, the pollen tube, sent
out immediately after pollination, penetrates to the archegonia
without interruption. See Appendix.

The pollen tube in Pinus begins to renew its penetration
of the nucellus during April, about a year after the mother
cell entered upon the reduction division, the large tube nu-
cleus enters the tube, where it is completely invested by starch
grains, and at the same time the generative cell divides into

CONIFERALES

93

the stalk cell and body cell (Fig. 69, L). In this case the two
cells, so far as we have observed or seen figured, are " fore
and aft " with reference to each other, and not side by side,
as in Cycads and Ginkgo, the stalk cell being nearer the old
spore wall. It appears, therefore, that the generative cell per-
sists for about eleven months without dividing. The pollen
tube branches as it traverses the nucellus, not so extensively as
do the tubes of Cycads and Ginkgo, but sufficiently to give evi-
dence of its primitive service as an absorbing or rhizoidal organ.
It consumes about two months in traversing the nucellus after
its second start, entering the archegonium about the 1st of July.

The liberation and descent of the body cell into the tube,
accompanied by the freed nucleus of the stalk cell whose wall
has been ruptured, and their association with the tube nucleus
near the tip of the tube, has been described in detail by Dixon
for Pinus silvestris, and his account applies as well to Pinus
Laricio, except that we find
no definite order of arrange-
ment in the relative posi-
tions of these three bodies.

Just before fertilization,
about two months after the
diA^ision of the generative
cell into stalk and body cells,
the latter divides and forms
the two male cells (mor-
phologically sperm mother
cells). It is this division
which in the Cycads is ac-
companied by the appear-
ance of blepharoplasts, and
results in the organization

of two ciliated male cells. So far as observed, however, bleph-
aroplasts do not appear in Conifers and the male cells do not
become ciliated. This is doubtless associated with the fact that
in Conifers the pollen tube has become a sperm carrier, while
in the Cycads it performs no such function. At the time of fer-
tilization, besides dense cytoplasm rich in starch, the tip of the
pollen tube contains four bodies, namely, tube nucleus, stalk
cell nucleus, and the two male cells (Figs. 70-72).

FIG. 72. — Juniperus Virginiana, showing
the ends of two pollen tubes ; in the tube
to the left the body cell has not yet di-
vided ; in the other the division into two
male cells has just taken place ; in both
cases the nuclei of the tube and stalk cells
are in front; x 160. — After STRASBURGER.

94:

MORPHOLOGY OF SPERM ATOPHYTES

The recorded variations from the above account are as fol-
lows : In the work upon Taxus, Juniperus, Cupressus, and Se-
quoia no mention is made of the two ephemeral vegetative cells.
They are so evanescent, however, that they may well escape the
notice of any observer who does not have a very close series and
good preparations. In Taxus Belajeff 30 observed that the body
cell divides unequally, the larger male cell functioning, and the
smaller one remaining in the tube during fertilization, appar-
ently incapable of functioning (Fig. 73). In this he was con-
firmed by Strasburger,31 and very recently by Jager.40 It is in-

Fio. 73. — Taxus baccata : A, male gametophyte, showing generative cell (a), tube nu-
cleus (*), and young tube (April 10th) ; B, later stage of same, showing stalk («) and
body (b) cells ; C", the large body cell passing into the tube, accompanied by the
nuclei of the tube, having divided into two very unequal male cells ; E, the larger
male cell discharged into the egg; A-C x 230, D and E x 180. — After BELAJEFF.

teresting to note that in Arnoldi's recent work on Cephalotaxus
Fortunei,43 a genus apparently very closely related to Taxus, the
two male cells are shown to be of the same size and apparent
vigor. In Sequoia Shaw 34 observed that the pollen -tube does
not penetrate the nucellus immediately, but advances across its
flat top and turns downward between it and the integument,
branching freely. Finally, one branch turns into the nucellus
about one fifth down its length, penetrates obliquely downward

CONIFERALES 95

and enlarges, and presently becomes indistinguishable among the
numerous tortuous sterile megaspores which cluster about the
upper end of the elongated fertile one. The branches of the
tube which remain external to the nucellus advance in every
direction between it and the integument. Such a condition of
things suggests the free branching rhizoidal habit of the tubes
of Cycads and Ginkgo, although in them the branching is within
the body of the nucellus. See Appendix.

FERTILIZATION

The tip of the pollen tube, having penetrated the overlying
tissue of the nucellus, reaches the wall of the embryo sac, and
either passes directly through it or flattens out upon it in a foot-
like expansion, sending out a small branch to pierce the wall.
In any event, the tube, or a branch of it, reaches the neck of the
archegonium, crushes the neck cells, and comes in contact with
the egg.

This contact may be regarded as the beginning of the pro-
cess of fertilization, which ends with the fusion of the sexual
nuclei. Jager 40 reports that in Taxus baccata the tube reaches
the egg about June 1st ; Coker 45 states that the date for Tax-
odium distichum is about June 15th ; the same date is given by
Arnoldi 43 for Cephalotaxus Fortunei\ while in Pinus Laricio
we find the date to be about July 1st. Of course the dates vary
somewhat with the season and with the latitude.

It has been shown by Blackman 37 and confirmed by Cham-
berlain 39 that the tip of the tube becomes fused with the egg
membrane, and does not enter the cytoplasm, as usually stated.
The appearance of an elongated and well-defined cavity in the
cytoplasm of the upper part of the egg is due to the inrush of the
contents of the tube (Figs. 74, 76, B).

The pit in the tip of the tube was described by Hofmeister,5
Strasburger,15 and others, but they claimed that it remains closed
by the primary wall membrane. In 1883 Goroschankin 26 stated
that he saw the pit perforated in Pinus Pumilio, which was
fully confirmed by Dixon 33 in Pinus silvestris, and by Black-
man 37 in the same species (Fig. 71). Nearly the whole of the
contents of the tube is injected into the cytoplasm of the egg.
In Pinus Pumilio , Goroschankin 26 observed both male cells pass
in, and Strasburger 28 confirmed it for Picea vulgaris. Dixon 33

96

MORPHOLOGY OP SPERM ATOPHYTES

seems to have been the first to see all the four structures in the
pollen tube injected into the egg, an observation which has been
frequently confirmed since. It should be remembered, however,
that in the case of Taxus baccata Belajeff 30 reports the passing
into the egg of a single male cell only, the smaller one re-
maining in the tube. In Pinus, in Taxodium as sho\vn by
Ooker,45 and in Cephalotaxus as shown by Arnoldi,43 the upper

part of the cytoplasm of the egg,
after the discharge of the tube, con-
tains two male cells, the disorgan-
izing tube nucleus and stalk cell nu-
cleus, besides cytoplasm of the tube
and starch grains. The cytoplasm of
the male cells is no longer distinguish-
able, and their nuclei may be recog-
nized by their large size and more
deeply staining contents as compared
with the tube and stalk cell nuclei,
which can hardly be distinguished
from each other.

One male nucleus moves forward
through the cytoplasm of the egg,
the other three injected bodies re-
maining near the top of the egg and
gradually disappearing. Arnoldi 43
makes an interesting observation in connection with the func-
tionless male cell in Cephalotaxus Fortunei. He reports that as
it lies unused in the peripheral region of the cytoplasm of the
egg it may divide amitotically, one of his figures showing six
nuclei as having been derived from a single functionless male
nucleus. According to Blackman,37 the functional male nu-
cleus moves very rapidly and increases in size, an increase appar-
ently due in some cases to actual increase of stainable substance,
and in others to vacuolation. The male nucleus of Pinus is
nearly spherical, with a diameter of about 40 /A, and is usually
one third (rarely one half) the diameter of the ellipsoidal or egg-
shaped nucleus of the egg.

The behavior of the male nucleus after it reaches the egg nu-
cleus is peculiar (Fig. 74). According to Blackman, it pushes in
the membrane of the egg nucleus, and finally comes to lie within

FIQ. 74. — Pinus silvestris, the
male nucleus entering the
female nucleus (June 19th) .
x 135. — After BLACKMAN.

COXIFERALES 97

it, both walls being intact. At this time similar changes occur in
the contents of both nuclei, and the spindle of the first segmenta-
tion begins to organize while the two nuclei are still distinct.
Blackman claims that no resting ferti-
lized nucleus is ever formed, although in -; ' ' ':„=»,.
all other known plants the male and fe- ;;•/•"' ^Sfe
male nuclei fuse while in the resting con- •- •£:
dition and form a definite resting ferti- .'•' >'*„„. ft
lized nucleus. Chamberlain confirms ; ; ^L^t *'\
this so far as to state that the two nuclei ™fe

"*"<§&•'•"• — — Q - - *,^

are in the spirem stage after the male '.,: ^/%M^^> *•'•
nucleus has entered the female nucleus V: T'
(Fig. 75). Woycicki 41 has also ob-
served in Larix Daliurica that in some .?'$
cases after the coalescence of the two :y: '•
nuclei two separate chromatin groups '' ^.^^ ^-''
are to be distinguished, one of which he Fie.76.--K»twZar*jfo,the

TI j ,1 -, j ,T ,-, chroraatinof the male and

would regard as the male, and the other female nuclei in the spi_
the female group. Such behavior finds rem stage within the lim-
its parallel among animals, as in As- its of the female nucleus ;

. „ i c x 500.— After CHAMBER-

cans, but not among plants, so far as LAIN
known. The process of fusion is exceed-
ingly slow, according to Blackman beginning when the male and
female nuclei begin to react upon one another, and closing
when the half chromosomes, derived from the male and female
nuclei respectively, fuse at the poles of the first segmentation
spindle. See Appendix.

Some minor deviations from this account of Blackman, given
for Finns, have been observed by Coker 45 in Taxodium dis-
t i chum, and by Jager 40 in Taxus baccata. In the former the
two nuclei come in contact near the upper end of the egg and
travel together toward the base, not fusing completely till the

* "In Van Beneden's epoch-making work on Ascaris it was shown not only
that the nuclei do not fuse, but that they give rise to two independent groups
of chromosomes which separately enter the equatorial plate and whose descend-
ants pass separately into the daughter nuclei. Later observations have given
the strongest reason to believe that, as far as the chromatin is concerned, a
true fusion of the nuclei never takes place during fertilization, and that the
paternal and maternal chromatin may remain separate and distinct in the
later stages of development — possibly throughout life." — Wilson's The Cell in
Development and Inheritance, second edition, p. 204

98 MORPHOLOGY OF SPERMATOPHYTES

bottom is nearly or quite reached. In Taxus there is the same
passing of the paired nuclei to the bottom of the egg for fusion,
and they are also reported as being of the same size. Perhaps
it is well to call attention to the fact that two free nuclei from
the early divisions of the fusion nucleus might be mistaken for
male and female nuclei unless the latter had been previously
observed and their great difference recognized. These free
nuclei are of the same size and pass to the bottom of the oospore.
It is also very possible that in some cases the ventral canal cell
is fertilized rather than the egg ; and there are some grounds for
believing that the nuclei of the ventral canal cell and of the egg
may fuse, suggesting what has been observed among animals in
cases of parthenogenesis (Wilson, loc. cit., pp. 280-283).

IV. THE EMBRYO

The first adequate account of the development of the embryo
in Conifers was that given by Strasburger in 18T2.17 Later
the same author 24 published more fully upon the subject, and
several investigators since have touched upon the embryogeny
of Conifers ; but the whole subject is in need of more detailed
investigation. While the earlier stages of the embryo are fairly
well known, our knowledge of the development of the embryo
proper is little more than an outline. We have obtained a com-
plete series of the early stages in Pinus Laricio, and from it the
following account is written.

1. Development of Proembryo. — The fusion nucleus retains
its central position in the cytoplasm of the oospore (Fig. 76, A\
and, beginning to divide, organizes the largest spindle of the
free nuclear series (Fig. 77), as in Cycads and Ginkgo. The
axis of this spindle is always inclined to the long axis of the
archegonium, and its fibers are not at all convergent, resulting
in a broad rectangular outline. The two daughter nuclei are
very small at first, but increase rapidly in size, and then
simultaneously divide. The four free nuclei which are
formed by this second division are at first very small, but
increase in size both through vacuolation and by taking in
substance (Fig. 76, B). When the four nuclei have reached
their full size, they begin to move toward the base of the spore,
and at this time show an investment of strong fibers, which

COXIFERALES

99

are directed tangentially. Chamberlain 39 thinks that this tan-
gential direction of fibers is caused by the rotation of the nuclei

';•-"

E

-•r

D

H

FIG. 76. — Pimis Laricio, early stage in the development of the embryo : A, fusion of
male and female nuclei; £, the first four nuclei of the proembryo, also numerous
nuclei from the jacket cells (/>, pollen tube; «, cavities caused by inrush of con-
tents of pollen tubes) ; C, the four nuclei at the base of the egg and dividing; />,
the first plate of four cells ; E, a somewhat later stage, showing partial cleavage of
the main mass of the egg; F, two plates of cells, the lower cells in process of divi-
sion ; G, the three tiers of cells of the completed proembryo, with the four free
nuclei above them (r, " rosette " tier ; », suspensor tier) ; H, the suspensor tier (s)
beginning to elongate; A- C collected June 25th, D-G July 2d; * 104.

100

MORPHOLOGY OF SPERMATOPHYTES

as they descend. Upon reaching the base of the oospore the four
nuclei arrange themselves in a plane and become ensheathed by
fibers derived from the nuclear membrane. Two walls are then
formed at right angles to each other and to the base of the spore,
each nucleus thus being separated from the others, but freely
exposed above to the general cytoplasm of the spore. Black-
man 37 speaks of " each nucleus thus being at the bottom of a
kind of shaft that opens above/' and the same author thinks that
the ensheathing fibers, which disappear at this time, have some
intimate connection with the formation of the walls. In any

event, the appearance of these
first incomplete walls is quite
apart from nuclear division,
and the incomplete segmenta-
tion of the cytoplasm presents
a peculiarity which is paral-
leled in the yolked eggs of
certain animals. In fact, the
method of receiving food sup-
ply into the egg, the organ-
ization of a small group of
cells to develop the embryo,
the setting aside of the great
mass of the oospore as a nutri-
tive supply, and the appear-
ance of the peculiar walls
about the basal nuclei seg-
menting more or less the cyto-
plasmic mass but not cutting
it off, are facts which are all more suggestive of the embryogeny
of certain animals than of other plants. Chamberlain observed
one case in which the first and second segmentations of the
fusion nucleus were accompanied by walls, a four-celled group
thus appearing near the center of the oospore.

According to our observation, the walls which separate the
four primitive nuclei, as mentioned above, do not appear until
after the nuclei are in an advanced stage of division (Fig. 76, C).
The spindles formed by the four basal nuclei are extremely
broad and multipolar, but the later spindles are definitely bipolar.
The basal nuclei divide simultaneously in a plane transverse

FIG. 77. — Pinus Laricio, the first division
of the nucleus of the oospore; the
entire mitotic figure is intranuclear;
x 500. — After CHAMBERLAIN.

COXIFL'RALES' 101

to the long axis of the oospore, a tier of four completely walled
cells thus being cut off below (Fig. 76, D), the upper nuclei
still being freely exposed above to the partially segmented
c\ tnplasni of the spore (Fig. 76, E), and increasing very much
in size. The cells of the single completely walled tier then
divide simultaneously, and two tiers are organized (Fig. 76, F).
The process is again repeated by the lower tier, and the result
is three tiers of four cells each (Fig. 76, G). AVith the appear-
ance of this last tier the development of the proembryo ceases.
At this stage, therefore, there are three complete tiers, with
four cells in each tier, and above these completely walled cells
there are the four nuclei separated from one another by walls,
but freely exposed above to the food supply of the spore. By
their size and generally active appearance it would seem that
they must continue functional for some time in connection with
nutritive work.

It will be noted that the proembryo in Conifers differs from
that in Cycads in the fact that free nuclear division is much
restricted, resulting usually in a group of four nuclei rather than
a great number ; that as a result there is no development of
tissue in a parietal position; that the tissue developed at the
base of the oospore is more definite and limited as to number of
cells and their functions. At the same time, the evidence seems
clear that the proembryo of Conifers must have been derived
from just such a condition as is shown by the proembryo of
( 'yoads and Ginkgo.

While this account probably applies in outline to Conifers
in general, there are differences in certain details which are
suggestive. The outline which seems to apply to all Conifers
is as follows : the development of free nuclei from the fusion
nucleus, the placing of these nuclei in a plane at the bottom of
the oospore, the appearance of incomplete walls separating the
nuclei from one another but permitting their free exposure to
the nutritive mass above, the development* from this plate of
nuclei of three tiers of completely walled cells, the development
of the embryo from the lowest tier of cells, the development of
the suspensor from the middle tier, and the retention of the
uppermost tier in the oospore.

The details which differ from those given in the above ac-
count for Pinus are as follows : Recently Arnold! 43 has de-

102 MbpfiOLCXrY^ OF 'SPERM ATOPHYTES

scribed a most interesting modification in the formation of the
proembryo in Cephalotaxus. The egg nucleus " divides three
or four times/' which means the production of eight or sixteen
free nuclei. These nuclei pass to the bottom of the oospore and
divide further, giving rise to " a number " of free cells, which
organize tiers " as in Taxus" Just how these numerous free
cells organize the tiers the author does not state, and his figures
indicate that the condition of his preparations would forbid any
claim beyond the mere existence of tiers. The number of free
nuclei and cells, however, before the organization of tissue, is
the important point, and indicates what would seem to be the
most primitive embryogeny among the Conifers. Although
normally in Pinus an embryo is developed at the end of each
suspensor cell — that is, a single oospore gives rise to four em-
bryos— we found in P. Laricio cases of a single embryo develop-
ing from the plate of four cells carried down by the four sus-
pensors. It seems that this is the normal behavior in Picea
excelsa, in which the suspensors remain coherent, forming a
single compound suspensor at the end of which a single embryo
develops. So far as the records go, the development of a single
embryo from the embryonal plate seems to be true in all groups
excepting Abieteae. Among the Cupresseae it is reported that
each of the three tiers of the proembryo is originally a single cell,
either remaining so, as in Cupressus, or dividing, with the ex-
ception of the embryonal cell, as in Tliuja, or all dividing, as
in Juniperus. If this sequence of events be true, it would seem
to follow that in these cases the fusion nucleus passes to the
bottom of the oospore without division, and there behaves in
building up tiers just as do the four basal nuclei in the case of
Pinus. Such statements, however, need further investigation.
The number of cells in each tier may also vary from four, the
usual number in Taxus being six, and sometimes reaching
even ten.

2. Development. of Suspensor. — The suspensor, or rather the
four suspensors, are formed by the remarkable elongation of the
cells which form the middle tier (Fig. 76, H). The lowest tier
is thus carried down into the midst of the endosperm, and the
uppermost tier remains in the base of the oospore, forming the
so-called " rosette " of Schacht. The suspensors become long
and tortuous (Fig. 78, A), and each bears at its tip one of the

COXIFERALES

103

four cells of the embryonal
tier, each of which begins the
development of an embryo.

In many Conifers the ap-
pearance of embryonal tubes
is a prominent feature. These
tubular processes are devel-
oped from basal cells of the
embryo, and seem to serve
as additional absorbing or-
gans (Fig. 78, B). They are
found investing the suspensor
and the base of the embryo,
and are sometimes exceeding-
ly numerous.

3. Development of Em-
bryo.— Concerning the devel-
opment of the embryo proper
our knowledge is meager.
In every case, with the excep-
tions noted below, it is devel-
oped from the lowest cell or
tier of cells at the base of the
oospore, the tier which is car-
ried down by the suspensor.
As has been stated, the cells
of this embryonal plate, usu-
ally four or six in number,
may together organize a sin-
gle embryo, or in Abieteae
each one is more apt to de-
velop separately. We ob-
served one case in Pinus La-
ricio which may serve to
indicate the possibilities in
connection with the devel-

FIG. 78. — Pinus Laricio* showing young
embryo : A. two young embryos with long
suspensors ; £, an older embryo, showing
three embryonal tubes.
* 8

104:

MORPHOLOGY OP SPERMATOPHYTES

opment of embryos. At the end of a single suspensor two
embryos had begun to develop, the first division of the em-
bryonal cell having been longitudinal, and the two daughter

cells having become so organically
dissociated that they pursued their
further history independently.
The embryo may thus start with
a plate of cells or with a single
cell, and hence there is probably
no very definite sequence of
events. In the case of embryos
which develop from a single cell,
the first two or three divisions are
usually transverse (Fig. 78), al-
though cases have been found in
which the first wall is vertical.
Whether there is any definite
sequence of events or not between
these early stages of the embryo
and the organization of its great
regions (Fig. 79) is uncertain, but
it is probable that the behavior
of these primitive cells may be
exceedingly variable, dependent
upon conditions which have not
yet been recognized.

The current statements as to
the structure of the root and stem
tips of Gymnosperms have been derived mainly from Stras-
burger.17 In reference to the root tip this author states that in
Gymnosperms in general the differentiation of meristem at the
apex of the root is essentially different from that of Angio-
sperms. A full statement of the difference was given under
the root. The differentiation of the regions at the apex of
the stem shows considerable variation in Gymnosperms. As
was remarked under the stem, in Dammara, Cunninghamia,
and certain species of Araucaria, the dermatogen, periblem,
and plerome are clearly distinguishable at the very apex; while
in the Abieteae the three regions merge in a common group
of initials, a character shared by the Cycads. It is interesting

FIG. 79. — Pinus Laricio, a fully de-
veloped embryo imbedded in the
endosperm.

CONIFERALES

105

to note that in Epliedra, even in the same species (E. campylo-
poda), both conditions are found.

Strasburger 24 records a remarkable embryonic structure in
Ccplicdotaxus Foriunei and Araucaria Brasiliana. The stem
tip, instead of being absolutely terminal, is covered by a small
group or layer of cells which is soon thrown off, but seems for
a time to serve as an organ of penetration and protection, finding
its analogue in the root cap (Figs. 80,
81). The presence of this terminal
group is reported to result in a shifting
of the functions of the tiers of the pro-
embryo (Fig. 80). The lowest tier,
which ordinarily develops the embryo,
in these cases organizes an embryo cap;
the middle tier, which ordinarily devel-
ops the suspensor, organizes the embryo;
the uppermost tier, or " rosette " of
Schacht, develops the suspensor; while
a fourth tier, apparently organized by
the uppermost nuclei which are exposed
above to the cytoplasm of the spore,
functions as the rosette. It is evident
that the statement sometimes made that
the stem tips of these peculiar embryos
are " endogenous " is not true in the
same sense that roots are endogenous.

The recorded numbers of cotyledons
are as follows: two in Taxaceae, in Cu-
presseae generally, and in some species
of Araucaria; three to five in some
Cupresseae; four to nine in Taxodieae;
and five to fifteen in Abieteae.

Seed and Fruit. — The origin of the
testa is but little known in detail, but
the region of its development may be
seen in Fig. 82. In Pinaceae it is hard
throughout, while in Taxaceae it is ac-
companied usually by a fleshy layer, or by a fleshy aril. In
Podocarpus and Cephalotaxus the seed covering consists of a
fleshy outer layer and an inner hard one, as in the seeds of Cycads

FIG. 80.—Cepkalotax'us For-
tunei : A, a proembryo still
within the oospore, show-
ing the cap of cells pro-
tecting the embryo proper
(shaded) ; B, a later stage,
in which the suspensor (s)
has developed; x 63. — Aft-
er STRASBURGER.

106

MORPHOLOGY OF SPERMATOPHYTES

and Ginkgo ; while in Taxus, Dacrydium, and Microcachrys the
hard coat is surrounded by a fleshy aril. The dry seeds are also
frequently winged, either by an outgrowth from the testa or by
flakes split from the ovuliferous scales.

Changes following fertilization also occur beyond the bound-
aries of the seed, and result in a more or less definite fruit.

FIG. 81. — Araucaria Brasiliana : A, a proembryo filling the entire oospore, only the
shaded cells belonging to the embryo proper, the wedge-shaped mass of cells
below being the embryo cap : B, a later stage, showing the embryo cap thrust
aside ; x 153. — After STRASBURGER.

The bracts or scales, or both, are affected, enlarging more or less
extensively, and becoming either woody, as in Abieteae, or
fleshy, as in Juniperus', while in Podocarpus sometimes the
whole strobilus becomes fleshy.

V. THE CLASSIFICATION OF CONIFERS

Our knowledge of the morphology and history of Conifers is
not sufficient to justify much certainty as to their classification.
Probably the scheme proposed by Eichler and Engler in Engler
and Prantl's " Natiirlichen Pflanzenfamilien " is the best expres-
sion of our present knowledge, at least it seems best adapted to
the purpose of this book. There is general agreement that

CONIFERALES

ior

there are at least two great divergent families of Conifers, the
Taxaceae and Pinaceae, differing from each other not only in
certain important morphological characters, but also in general
geographic distribution.

The Taxaceae are characteristically south temperate forms,
and include only about seventy-one species. Generally true
strobili are not organized, at least they do not conceal the ovules,
which are freely exposed. This would seem to diminish the
chances of pollination as compared with the arrangement in
Pinaceae, where the scales are of some service in catching pollen.
Moreover, the seeds always develop a partially fleshy testa or
an aril, which seems to be corre-
lated with their exposed con-
dition.

The Pinaceae, on the other
hand, are characteristically north
temperate forms, and include
about two hundred and twenty-
four of the two hundred and
ninety-five known species of Con-
ifers, and in individual or forest
display the difference is much
greater. Some explanation of
this more favorable adaptation
to present conditions may be
found in the fact that the organ-
ization of true strobili not mere-
ly protects the ovules, but also
increases the chances for pollina-
tion. In the maturing of the
seed the testa is woody or leath-
ery, with no fleshy development,
such as is characteristic of Cycads
and Ginkgo, and of the Taxaceae,
not even in the form of a fleshy

aril. AVith such characters, and also with the northern distribu-
tion, the Pinaceae stand out quite distinctly.

The Taxaceae, with their exposed ovules and succulent seed
investment, contain but eight of the thirty-five genera of Coni-
fers, and only seventy-one of the two hundred and ninety-five

FIG. 82. — Pinus Laricio, longitudinal
section of the ovule showing1 the be-
ginning of the testa (shaded zone) :
•£, integument ; p, pollen tubes ; «-,
nucellus ; a, archegoniuni ; y, endo-
sperm ; £, testa ; x 14.

108 MORPHOLOGY OF SPERMATOPHYTES

species. Of these seventy-one species, forty belong to the single
genus Podocarpus, as characteristic of the south temperate re-
gions as Pinus is of the north temperates. At least two series
of Taxaceae are evident, the Podocarpeae and the Taxeae. In
the former, containing fifty-four species, the ovules are appar-
ently foliar and the pollen grains are winged ; while in the lat-
ter, containing seventeen species, the ovules are apparently
cauline and the pollen grains are wingless.

The Pinaceae display at least four distinct series of forms, in
three of which (Araucarieae, Abieteae, and Taxodieae) the spiral
leaf arrangement prevails throughout, in bracts and sporophylls
as well as in the ordinary leaves; and in one (Cupresseae) the
leaf arrangement is cyclic throughout. The Araucarieae, con-
taining about fourteen species, and prominent in the southern
hemisphere, have almost completely coalescent bract and scale,
solitary ovules, and free pendulous pollen sacs. The Abieteae,
containing about one hundred and twenty-two species and form-
ing the prominent coniferous forest display of the northern
hemisphere, have almost completely distinct bract and scale,
winged pollen, and needle leaves. The Taxodieae, containing
about twelve species, have almost completely coalescent bract
and scale, and wingless pollen. The Cupresseae, containing
seventy-eight species, are distinguished from the other groups
of Pinaceae by the cyclic character, and may have the strobili
ripening hard or fleshy (as in Juniperus).

The morphological evidence seems to point to the belief that
the Taxaceae represent the more primitive forms of Conifers,
and that among the more recent Pinaceae the Abieteae, and
notably the genus Pinus, stand for the most specialized forms.

A tabulated statement of the classification outlined above is
as follows, the numbers following each genus indicating the
approximate number of recognized species it contains :

A. TAXACEAE
I. Podocarpeae

1. Saxogothaea 1

2. Microcachrys 1

3. Podocarpus 40

4. Dacrydium 12

II. Taxeae

5. Phyllocladus 3

6. Cephalotaxus 4

7. Torreya 4

8. Taxus . . 6

CONIFERALES

109

9. Agathis

B. PINACEAE
I. Araucarieae
4 | 10. Araucaria

10

II. Abieteae

11. Firms 70

12. Cedrus 3

13. Larix 8

14. Pseudolarix 1

15. Picea.. . 12

16. Tsuga 4

17. Pseudotsuga 2

18. Keteleeria 2

19. Abies.. . 20

20. Sciadopitys

21. Cunninghamia 1

22. Sequoia 2

23. Arthrotaxis . . 3

III. Taxodieae
1

24. Cryptomeria 1

25. Taxodium 2

26. Glyptostrobus 2

IV. Cupresseae

27. Actinostrobus 2

28. Callitris 15

29. Fitzroya 2

30. Thujopsis 1

31. Libocedrus. . 8

32. Thuja 4

33. Cupressus 12

34. Chamaecyparis. . , 4

35. Juniperus 30

LITEKATURE CITED

1. BROWN, ROBERT. Character and Description of Kingia, a new

Genus of Plants found on the S. W. Coast of New Holland, with
Observations on the Structure of its Un impregnated Ovulum and
the Female Flower in Cycadeae and Coniferae, 1827. Included
also in Brown's Miscellaneous Botanical Works. 1 : 1866.

2. SCHLEIDEN, M. J. Sur la Signification Morphologique du Placen-

taire. Ann. Sci. Nat. II. 12: 373-376. 1839.

3. VON MOHL, HUGO. Ueber die Mannlichen Bliithen der Coniferen.

Verm. hot. Schriften 45-61. 1845. (Published as a dissertation,
1837.)

4. BRAUN, A. 1853.

5. HOFMEISTER, W. Embryobildung der Phanerogamen. Jahrb. f.

wiss. Bot. 1 : 82-188. pis. 7-10. 1858.

6. BAILLON, H. Recherches organogeniques sur la fleur femelle des

Coniferes. Ann. Sci. Nat. IV. 14: 186-199. pis. 12-13. 1860.

7. CASPARY, R. De Abietinearum floris feminei structura morpholo-

gica. Ann. Sci. Nat. IV. 14: 200-209. 1860.

8. DICKSON, A. Observations on some Bisexual Cones occurring in

the Spruce Fir (Abies excelsa). Trans. Edinburgh Bot. Sci. 6:
418-422. 1860.

HO MORPHOLOGY OF SPERMATOPHYTES

9. HOFMEISTER, W. Vergleichende Untersuchungen. English trans-
lation. 1862.

10. OERSTED, A. S. Bidrag til Naaletroeernes Morphologi. Videnskab.

Meddelelser Nat. Foren. Copenhagen. 1864.

11. PARLATORE, F. Studi Organografici sui Fiori e sui Frutti delle

Conifere. Opuscula botanica. 1864.

12. HOFMEISTER, W. Lehre der Pflanzenzelle. 1867.

13. SACHS, J. Lehrbuch. 1868.

14. VAN TIEGHEM, PH. Anatomie comparee de la fleur femelle et du

fruit des Cycadees, des Coniferes, et des Gnetacees. Ann. Sci.
Nat. V. 10: 269-304. pis. 13-16. 1869.

15. STRASBURGER, E. Die Befruchtung bei den Coniferen. 1869.

16. VON MOHL, HUGO. Morphologische Betrachtung der Blatter von

Sciadopitys. Bot. Zeit. 29: 17-23. 1871.

17. STRASBURGER, E. Die Coniferen und Gnetaceen. 1872.

18. EICHLER, A. W. Sind die Coniferen Gymnosperm oder nicht ?

Flora 54: 241-247, 260-272. 1873 ; and later papers.

19. ENGELMANN, G. Morphology of the Carpellary Scales of Coniferae.

Am. Jour. Sci. III. 12: 469-470. 1876.

20. STENZEL, G. Beobachtungen an durchwachsenen Fichtenzapfen.

Ein Beitrag zur Morphologic der Nadelholzer. Nov. Act. Nat.
Cur. 38: 289-350. pis. 12-15. 1876.

21. STRASBURGER, E. Die Befruchtung und Zelltheilung. Jen.

Zeitsch. 2: 435-536. pis. 27-35. 1877.

22. STRASBURGER, E. Befruchtung und Zelltheilung. 1878.

23. CELAKOVSKY, L. Zur Gymnospermie der Coniferen. Flora 62 :

257-264, 272-283. 1879.

24. STRASBURGER, E. Die Angiospermen und die Gymnospermen.

v 1879.

25. CELAKOVSKY, L. Zur Kritik der Ansichten von der Fruchtschuppe

der Abietineen. Abhandl. d. konigl. bohm. Gesell. d. Wiss. VI.
11: 1882.

26. GOROSCHANKIN, J. Zur Kenntniss der Corpuscula bei den Gym-

nospermen. Bot. Zeit. 41 : 825-831. pi. 7 A. 1883.
Ueber den Befruchtung-Process bei Pinus Pumilio. 1883.

27. DE BARY, A. Comparative Anatomy of the Vegetative Organs of

the Phanerogams and Ferns. English translation. 1884.

28. STRASBURGER, E. Neue Untersuchungen, etc. 1884.

29. CELAKOVSKY, L. Die Gymnospermen : eine morphologisch-phy-

logenetische Studie. Abhandl. d. konigl. bohm. Gesell. d. Wiss.
VIII. 4: 1890.

30. BELAJEFF, W. Zur Lehre von dem Pollenschlauche der Gymno-

spermen. Ber. d. deutsch. bot. Gesell. 9: 280-286. pi 18. 1891.

31. STRASBURGER, E. Ueber das Verhalten des Pollens und die

Befruchtungsvorgange bei den Gymnospermen. Hist. Beitr. 4 :
1-158. pis. 1-3. 1892.

COXIFERALES HI

31a. MOTTIER, D. M. On the Archegonium and Apical Growth of the
stem in Tsuga Canadensis and Pinus silvestris. Bot. Gaz. 17:
141-143. pi. 8. 1892.

32. BELAJEFF, W. Zur Lehre von dem Pollenschlauche der Gym-

nospermen (Zweite Mittheilung). Ber. d. deutsch. hot. Gesell.
11 : 196-201. pi. 12. 1893.

33. DIXON, H. N. Fertilization of Pinus silvestris. Ann. Bot. 8 : 21-

34. pis. 3-5. 1894.

34. SHAW, W. R. Contribution to the Life-history of Sequoia. Bot.
v Gaz. 21: 332-339. pi. 24. 1896.

35. CELAKOVSKY, L. Nachtrag zu meiner Schrift iiber die Gymno-

spermen. Engler's Bot. Jahrb. 24 : 202-231. 1898.

36. COULTER, J. M. Notes on the Fertilization and Embryogeny of

Conifers. Bot. Gaz. 23: 40-43. pi. 6. 1897.

37. BLACKMAN, V. H. On the Cytological Features of Fertilization

and Related Phenomena in Pinus silvestris L. Phil. Trans.
Roy. Soc. 190: 395-426. pis. 12-14. 1898.

38. CHAMBERLAIN, C. J. Winter Characters of certain Sporangia.

Bot. Gaz. 25: 125-128. pi. 11. 1898.

39. CHAMBERLAIN, C. J. Oogenesis in Pinus Laricio. Bot. Gaz. 27 :

268-280. pis 4-6. 1899.

40. JAGER, L. Beitrage zur Kenntniss der Endospermbildung und

zur Embryologie von Taxus baccata. Flora 86 : 241-288. pis.
15-19. 1899.

41. WOYCICKI, Z. On Fertilization in Coniferae (in Russian), pp. 1-57.

pis. 2. 1899.

42. ARNOLD, W. Bull. Soc. Imp. Nat. Moscou. 329-341. pis. 2. 1899

(1900) ; see review in Jour. Roy. Micr. Soc. 1900 : 428.

43. ARNOLDI, W. Beitrage zur Morphologic der Gymnospermen. III.

Embryogenie von Cephalotaxus Fortunei. Flora 87: 46-63.
pis. 1-3. 1900.

44. ARNOLDI, W. Beitrage zur Morphologic der Gymnospermen. IV.

Was sind die " Keimblaschen " oder " HofmeisterVKorperchen "
in der Eizelle der Abietineen ? Flora 87 : 194-204. pis. 6-7.
1900.

45. COKER, W. C. Observations on the Gametophyte and Embryo of

Taxodium distichum. Johns Hopkins Univ. Circ. 19: 45-46.
1900.

46. GOEBEL, K. Organography of Plants. English translation. 153-

155. 1900.

47. WORSDELL, W. C. The Structure of the Female " Flower" in

Coniferae. Ann. Bot. 14: 39-82. 1900.

48. WORSDELL, W. C. The Vascular Structure of the Ovule of

Cephalotaxus. Ann. Bot. 14: 317-318. 1900.

49. MURRILL, W. A. The Development of the Archegonium and Fer-

tilization in the Hemlock Spruce (Tsuga Canadensis). Ann.
Bot. 14: 583-607. pis. 31-32. 1900.

CHAPTEE IV

GNETALES

THIS group includes three genera which differ remarkably
in habit and habitat. Certain angiospermous characters which
they display have suggested that they may have given rise to
Angiosperms, but such a theory seems to have been abandoned
by most morphologists. The two characters of the group which
have been regarded as distinguishing it from other Gymno-
sperms are (1) the occurrence of "true vessels" in the sec-
ondary wood, and (2) the presence of a " perianth." A mere
statement of these characters is apt to lead to an exaggerated
conception concerning their importance. The " true vessels "
are usually intermediate in character between the tracheids of
ordinary Gymnosperms and the tracheae of Angiosperms, and
are as easily associated with the former as with the latter. The
" perianth " is a doubtful structure, and is probably no more the
equivalent of the perianth of Angiosperms than 'are the bracts
closely related to the stamens and ovules in other Gymnosperms.

In addition to the two distinguishing characters given
above, the group has the four following characters in common,
but not peculiar to it : ( 1 ) opposite leaves ( a cyclic character
also displayed by the Cupresseae); (2) dicotyledonous embryos
(found also in Cycads and in several groups of Conifers) ; (3)
cauline ovules (true also of Taxaceae and perhaps of all Coni-
fers) ; and (4) no resin ducts (true also of Taxus). To these
characters perhaps should be added the remarkable prolonga-
tion of the integument of the ovule to form a long tubular
micropyle.

The genera are Ephedra, with about thirty species, from the
arid regions of both hemispheres; the monotypic Tumboa
(Welwitschia), from the extremely arid regions of western
112

GNETALES 113

South Africa ; and Gnetum , with about fifteen species, from the
tropics of both hemispheres.

The important morphological literature of the group may
be outlined historically as follows: In 1863 Hooker1 published
an account of Tumboa in an elaborate memoir, which is probably
more generally accessible in Eichler's 2 very full review. In
1872 Strasburger 3 published the first adequate account of the
more recondite morphological features of the group. In 1877
Beccari 4 gave an account of the ovulate flowers of Gnetum, and
discussed the nature of the floral structures; while in 1879
Strasburger 5 extended and . co-ordinated his previous observa-
tions. It is chiefly from the two publications of Strasburger that
the current accounts of text-books have been derived. In 1881
Bower 6' 7 published accounts of the embryo of Tumboa ; and in
1882 the same author 8 performed a similar service in the case of
Gnetum. These two accounts still remain practically the only
source of information concerning the embryos of Tumboa and
Gnetum. In 1892 Strasburger 9 again discussed the Gnetales,
and in the same year Karsten 10 published a preliminary paper
upon the genus Gnetum. In 1893 Karsten ll> 12 published two
additional papers, in which he modified and extended his pre-
vious conclusions. In 1894 and 1895 Jaccard 13' 14 published
accounts of Epliedra Helvetica, which assisted materially to an
understanding of the group. The most recent contribution is
that of Lotsy,15 who has given a very important account of the
female gametophyte and fertilization in Gnetum Gnemon. He
has also done great service in bringing together the essential
features of the scattered literature, and in contrasting the re-
sults of different investigators.

In the following account of Gnetales these variojus authors
have been laid under contribution, as we have had no oppor-
tunity to examine any properly fixed material.

I. THE VEGETATIVE ORGANS

The species of Ephedra are low straggling shrubs, with long-
jointed and fluted green steins, and opposite scalelike leaves con-
nate into a two-toothed sheath. True foliage leaves are gen-
erally lacking, the work of photosynthesis being assumed by the
stem, as in the case of the Equisetales, whose stems are very sug-
gestive of Ephedra (Fig. 83).

MORPHOLOGY OF SPERMATOPHYTE3

Tumboa is a remarkable monotypic genus. As described
and figured by Hooker,1 the body has the shape of a gigantic
radish, which rises little above the surface of the ground, and
whose crown is sometimes 3.5 to 4.5 meters in circumference.

FIG. 83. — Ephedra antisypTiilitica : J, staminate branches ; 2, ovulate branches ; <?, sterile
branches; 4-, staminate strobilus; 5, ovulate strobilus ; 6, longitudinal section of
seed ; 7, bracts of the ovulate strobilus. — After WATSON.

The broad top is more or less concave and somewhat two-lobed,
and from the deep horizontal groove which separates the crown
from the body two enormously long parallel-veined leaves arise,
which extend upon the ground sometimes for 3 to 3.5 meters,

GXETALES

115

and become split into numerous ribbons (Fig. 84). This single
pair of opposite leaves succeeds the cotyledons, and is the only
pair of foliage leaves produced, growing continually at the base,

where it is protected by the groove. These leaves last through
the lifetime of the plant, which is said to reach more than a hun-
dred years.

The species of Gnetum are either small trees or woody climb-

116 MORPHOLOGY OF SPERMA.TOPHYTES

ers, being among the prominent lianas of tropical forests (Fig.
85). The foliage is leathery in texture, and is very suggestive
of Dicotyledons, the well-developed opposite leaves being lan-
ceolate to ovate in outline and pinnately net-veined (Fig. 88).

THE STEM

Ephedra. — The characteristic long-jointed and fluted green
stems have been spoken of above. In the stem tip the dermato-
gen and periblem cap the plerome cylinder with distinct layers,
each layer having its own initials, or the three embryonic regions
merge in a common group of initials. It is of interest to note
that both of these conditions have been found in a single species
(E. campylopoda) . There are no cauline bundles, as in Turn-
boa and Gnetum, and no secondary cambium. The primary
cambium develops a certain amount of secondary tissue, but the
increase in diameter is never striking. The tracheary vessels
of the secondary wood, which distinguish Gnetales from other
Gymnosperms, are found in Ephedra at the inner part of the
secondary xylem cylinder, and consist of broad vessels associated
with ordinary gymnospermous tracheids. These vessels are in-
terrupted, moreover, by oblique walls, and display bordered pits
as well as simple pits. In every way they seem to be modified
gymnospermous tracheids. A peculiar feature of the genus is
the diaphragmlike plate of cells which occurs at the base of each
internode, rendering the stem easily separable at the nodes.
The stomata occur upon the fluted stem in rows, as in Equisetum.
and are sunken in urn-shaped depressions formed by mounds
of cuticle.

Tumboa. — The anatomy of the stem is somewhat confusing
on account of the extremely shortened axis. In the region
where the short crown and large tap root merge there is a broad
stratum of collateral bundles, which De Bary recognizes as two
layers very close together, and calls the " layer of the leaf trace."
The upper layer connects with the numerous veins of the leaves,
and the lower layer is associated with the system of vessels in
the tap root. The primary cambium is short-lived, and a series
of irregularly concentric bundles suggests a succession of sec-
ondary cortical cambiums. The resemblance to Cycads is fur-
ther marked by the occurrence of cauline bundles, not only in
the cortex but also in the pith. It is reported also that tracheids

FIG. 85. — Gnetum oval/folium, climbing on Cocos nucifera. — After KARSTEX.

MORPHOLOGY OF SPERM ATOPHYTES

with bordered pits are entirely lacking, simple pitted and spiral
vessels replacing them. In connection with the vascular bun-
dles there is a conspicuous development of fibrous sclerenchyma.
In the stem as a whole parenchymatous tissue largely predomi-
nates, and forms a conspicuous water-reservoir system.

Perhaps the most remarkable anatomical feature is the occur-
rence of the so-called " spicular " cells, found in large numbers
and scattered throughout all organs. They are very large, fusi-
form or branched, straight or variously curved, with greatly
thickened walls, in the outer layers of which many finely formed
crystals of calcium oxalate are imbedded close together.

Gnetum. — As in Tumboa, the primary cambium is short-
lived, and successive cortical cambium zones organize concentric
series of cauline bundles. As in Ephedra, the xylem contains
vessels of the true tracheary type associated with tracheids of the
Gymnosperm type.

THE LEAF

Ephedra. — The opposite leaves of Ephedra are reduced to
scales, which coalesce to form at each node a two-toothed sheath.
The occurrence of true foliage leaves seems to be- extremely rare.

Tumboa. — In the single pair of long-lived parallel-veined
leaves the stomata are in rows and deeply sunken. The chloro-
phyll-bearing tissue beneath the epidermis ensheaths a promi-
nent central tissue which is colorless or nearly so, and consti-
tutes a conspicuous water reservoir. The colorless middle tis-
sue is traversed by the very numerous and strong parallel
bundles, which are connected by transverse branches. These
branches sometimes end freely in the mesophyll, or send out
branchlets which end blindly. "While the general system of vena-
tion is of the monocotyledonous type, the blind ending of certain
branches is a dicotyledonous feature.

Gnetum. — The leaves of Gnetum seem to differ in no essen-
tial feature from such leaves of Dicotyledons as are leathery in
texture, having a typical reticulate bundle system of the pinnate
variety.

THE ROOT

The very meager accounts of the roots of Gnetales have refer-
ence only to species of Ephedra. From these it appears that
the group possesses the same peculiarities shown by the other

GXETALES

119

Gymnosperms. The root cap is derived in the same way from
the outer exfoliating layers of the periblem, and the primary
vascular cylinder is diarch.

II. THE SPOKE-PRODUCING MEMBERS

THE MICEOSPOEANGIUM

So far as we can discover, there is no recorded information
as to the development of the microsporangium in Gnetales, but
presumably it follows the general sequence common to Gymno-
sperms and Angiosperms. The following account, therefore,
can deal only with the gross structure of the so-called staminate
flowers.

Ephedra. — The flowers are monosporangiate, and are borne
in the axils of bracts in a spicate inflorescence. The staminate

FIG. 86. — Tumboa', A, young staminate flower; .Z?, older staminate flower, with the
.sterile ovule in the center : f, two stamens ; Z>, the sterile ovule ; E, the ovulate
flower, showing tubular micropyle (i) and winglike expansions of the " perianth " ;
F. the same after fertilization ; G, longitudinal section of a ripe seed, showing con-
spicuous suspensor (s). — After STRASBURGER, the figure being taken from ENGLER
and PRANTL'S Xat. Pflanzenfam.

flower consists of two more or less connate scales (" perianth ")
investing a projecting axis, which bears two or more sporangia.
The sporangia appear to be cauline, the sporangiferous axis re-
maining simple and bearing two sporangia, or branching some-
what above and bearing several sporangia (Fig. 83).
9

120

MORPHOLOGY OF SPERMATOPHYTES

Tumboa. — While the flowers are functionally monosporan-
giate, there is evidence of their derivation from a bisporan-

giate condition. They are borne
in the axils of prominent opposite
bracts which form conelike stro-
bili. These strobili are produced
upon branching axes, which arise
from the crown above the foliage
leaves, or rarely beneath them.
The staminate flower consists of
two decussate pairs of free bracts,
forming the so-called perianth,
within which there is a whorl of
six (rarely five or four) monadel-
phous stamens, each of which is
capped by a group of three spo-
rangia. In the center is a single
sterile ovule, whose projecting
micropyle is spirally coiled and
has a flaring expansion at the tip
(Figs. 86, 87).* The whole struc-
ture is puzzling and anomalous
among Gymnosperms, for it seems
to indicate derivation from a per-
fect flower. The microsporangia also seem to be distinctly
foliar in origin.

Gnetum. — The flowers are monosporangiate and mostly
dioecious, the inflorescence being in the form of spikes of con-
nate bracts with numerous flowers in their axils. The stami-
nate flowers resemble those of Ephedra, consisting of two coher-
ent bracts investing a short axis which bears two sporangia.
Such evidence as we have indicates that these sporangia are
cauline (Fig. 88).

THE MEGASPORANGIUM

Ephedra. — In the spicate inflorescence one or more ovulate
flowers occur in the axil of each bract. Each flower consists of
two more or less connate scales which invest a solitary cauline

FIG. 87. — Tumboa : A, staminate stro-
bilus ; J3, ovulate strobilus ; both
two thirds natural size. — A from
the Bot. Mag., B after LE MAOUT
and DECAISNE ; the whole figure
from ENGLER and PKANTL'S Nat.
Pflanzenfam.

* Excellent illustrations of the staminate and ovulate strobili may be found
in Gardeners' Chronicle III. 24 : 63. 1898.

GNETALES

121

ovule whose integument is produced into a very long tubular
projecting micropyle. In fruit the bracts are much modified,
becoming red and succulent, or much enlarged and chaffy, and
the " perianth " persists and becomes lignified (Fig. 83).

The development of the megasporangium, as given by Stras-
burger 3 for E. campyloda, indicates the usual sequence of
events. The archesporium consists of a group of hypodermal
cells, which by periclinal division cut off wall cells. The ex-
tensive periclinal division of the wall cells organizes them into
very regular longitudinal rows (Fig. 89, A), and places the pri-
mary sporogenous cells deep within the nucellus. Whether the
primary sporogenous cells divide or pass directly into mother
cells is not clear, but in any event the mother cells form a deep-
seated group variable in number or even reduced to one. x

Turnboa. — In the conelike strobilus the ovulate flowers are
axillary, and contain a single ovule, with no trace of stamens.

FIG. 88. — Gnetum : A, branch with leaves and staminate strobili ; B, upper portion of
a staminate strobilus ; <7, a staminate flower ; Z>, a branch with ovulate strobili ;
E, upper portion of an ovulate strobilus ; F, an ovulate flower ; G, longitudinal sec-
tion of an ovulate flower, showing the three envelopes (ii,ie,p}\ A-E, Gnetum
latifolia; F-G, Gnetum sp.t—A~E after BLUME, F and G after STRASBURGEB :
whole figure from ENGLER and PRANTI/S Nat. Pflanzenfam.

The perianth consists of a single pair of completely united
bracts, which close over the ovule like an outer integument and
become much flattened laterally, forming two opposite winglike
expansions. The integument directly investing the ovule forms

122 MORPHOLOGY OF SPERMATOPHYTES

the usual long projecting tubular micropyle, which is neither
spirally coiled nor expanded at the tip as in the sterile ovule of
the staminate flower. In fruit the enlarged bracts become
bright scarlet, and the seed is winged by the expanded perianth
(Figs. 86, 87).

Very little is known concerning the details of the develop-
ment of the megasporangium beyond the fact that a group of
mother cells is deep-seated beneath long rows of wall cells, as in
other Gymnosperms. The rows of wall cells are organized into
a somewhat persistent nucellar cap or beak, recalling the Cyca-
dales and Ginkgoales. There is no evidence that the beak devel-
ops a pollen chamber, but it is said to become riddled with pas-
sageways which are utilized by the ascending archegonial tubes
and descending pollen tubes.

Gnetum. — According to Lotsy,15 in G. Gnemon the very
young ovulate spike consists of a series of close-set cups, each
cup representing coalescent bracts. Later the cups are well
separated by the lengthening of the internodes, and within each
a prominent ringlike cushion appears. From the lower por-
tion of the cushion abundant yellowish paraphysislike hairs are
developed, while from the top of the cushion the more or less
numerous ovulate flowers arise, each flower apparently repre-
senting an axillary bud (Figs. 88, E, and 91, A). The solitary
ovule is organized directly from the tip of the bud axis, and
hence is cauline, while about it there arise in acropetal succes-
sion three envelops, the homologies of which have given rise
to considerable discussion (Fig. 91, G). The outermost envelop
is green and fleshy, resembling the perianth in Epliedra^ the
middle one is more delicate, not green, and does not rise so high
as the outer one; while the innermost is the long-necked struc-
ture called an integument in Ephedra and Tuniboa. The tip of
the micropylar tube is more or less lacerate, and the lobes spread
during pollination (Fig. 88, F).

Strasburger 9 discovered that there are two kinds of ovulate
flowers, those of the ovulate spikes, described above, and those
found in the staminate spikes. The latter differ in having but
two envelops about the ovule, the middle one of the complete
flower being absent. These incomplete flowers do not function,
but sometimes they develop far enough to contain embryo sacs.
In the very rare cases when an ovulate flower on a staminate

GNETALES

123

spike functions, Lotsy 15 observed that it possesses the three en-
velops of the complete ovulate flowers. Karsten 10j n* 12 found
that the middle envelop appears during the development of the
incomplete ovulate flowers, but gradually disappears.

Strasburger 9 regards all three envelops as integuments, the
middle and innermost corresponding to the inner envelop in
Ephedra. Beccari 4 first described the organogeny of the
flower, discovering that the three envelops appear in acropetal
succession. This seemed to militate against the theory of ovular
integuments, which appear in basipetal succession. Beccari's
interpretation, however, seems somewhat fanciful, since he re-
gards the middle envelop as the homologue of the staminate
cycle in Tumboa, the outermost envelop being a perianth.
Lotsy regards the two outer envelops as two whorls of bracts,
which he calls the perianth, and claims that there is only a single
integument. He lays stress upon the fact that in the two outer
envelops there is vascular tissue, claiming that this indicates
their foliar nature.

If Celakovsky's general claim be accepted, that there are
two integuments in all Gymnosperms, the so-called perianth of
Gnetales becomes an outer integument, and the middle envelop
of the ovulate flowers of Gnetum may well be only an outer layer
of the inner integument, as Strasburger has suggested. In
any event, there seems to be no question as to the tegumental
nature of the innermost envelop, and the acropetal succession
of the envelops would indicate that the one or two outer ones
are modified bracts. It must be remembered that this does not
necessarily militate against Celakovsky's view, since he regards
integuments as of foliar origin.

The development of the megasporangium has been described
in detail by Strasburger 3 for Gnetum Gnemon, and coincides
with the general sequence in Gymnosperms. The archesporium
consists of a group of hypodermal cells which cut off wall cells
by periclinal division. By repeated periclinal divisions the
wall cells form the usual thick mass of sterile tissue above the
sporogenous region (Fig. 91, F). Before pollination the cells at
the tip of the nucellus disorganize, an irregular pollen chamber
being the result, much as in Pinus. The several primary spo-
rogenous cells divide more or less and organize the deep-seated
group of mother cells.

MORPHOLOGY OF SPEKMATOPIIYTES

III. THE GAMETOPHYTES

THE FEMALE GAMETOPHYTE

While the female gametophyte conforms in a general way to
the Gymnosperm type, there are certain remarkable deviations
in Tumboa and Gnetum which are suggestive of a possible
method of derivation of the embryo sac structures of Angio-
sperms.

B

FIG. 89. — Ephedra : A, young ovule showing a row of three potential megaspores ; B,
upper part of nucellus after fertilization, showing the free cells of the proembryo
within the two oospores, x 64 ; f, suspensors developing from the free cells, passing
from the oospore into the endosperm, and each bearing at tip an embryonal cell,
x 64; D-F, later stages in the development of the embryo, x 154; A, E. campy-
lopoda ; B-F, E. altissima. — After STRASBURGER.

Ephedra. — According to Strasburger's account 3 for E. cam-
pylopoda, the mother cells organize a row of three potential
megaspores, the lowest of which is the fertile one (Fig. 89, A).
In case several fertile megaspores begin to enlarge, one finally
dominates and develops the gametophyte.

The details of the germination of the megaspore are lacking,
but Strasburger's outline is the usual one among Gymnosperms.

GXETALES

125

'he nucleus of the spore begins free nuclear division, the nuclei
become imbedded in a parietal layer of cytoplasm which sur-
rounds a great central vacuole, and free nuclear division con-
tinues until numerous nuclei appear within the enlarging em-
bryo sac. Walls next appear, and a layer of parietal tissue
develops, which by continued cell division gradually fills the
sac with a compact tissue, as in Conifers. The archegonia have
unusually long necks, and apparently a remarkably persistent
ventral canal cell.

It is evident that in Ephedra there is the least departure
from what may be called the Gymnosperm type of female
gametophyte, and presumably Ephedra may be regarded as the
most primitive of the Gnetales.

Tumboa. — This genus needs re investigation from the stand-
point of modern technique and modern morphology, for the
facts recorded concerning it are somewhat vague and may be
misleading. Xothing is known of the development of the soli-
tary functional megaspore further than that it lies at first deep
within the nucellus beneath long rows of wall cells. As this is
true of Gymnosperms in general, it may be assumed that the
history up to this point is the same.

Apparently the germination of the megaspore also proceeds
in the usual way, with free nuclear division, imbedding of the
free nuclei in a parietal layer of cytoplasm, appearance of walls,
and gradual centripetal filling of the embryo sac. It is evident,
however, that the tissue in the micropylar end of the embryo
sac is less compacted than that in the antipodal end. In any
event, the archegonium initials upon the micropylar surface of
the gametophyte behave in an exceedingly peculiar and inde-
pendent fashion. It is reported that there are from twenty to
sixty of these initials, and that they proceed no further in the
development of archegonia than to assume the appearance of
ordinary archegonium initials, an egg being organized directly
within each of them without any division. Just why the devel-
opment of archegonia should stop with the selection and en-
largement of initial cells is hard to understand except upon the
hypothesis that they are practically free cells. Each of these
cells, representing a greatly reduced archegonium, contains an
«gg which is ready for fertilization. There is suggested an inter-
mediate condition between Epliedra, with its complete arche-

126

MORPHOLOGY OP SPERM ATOPHYTES

gonia, and Gnetum, with its free eggs. No less remarkable than
the structure of the archegonium, if such it may be called, is its
behavior in connection with fertilization. It is reported that
each archegonial cell develops a long tubular process which
penetrates into the sterile tissue of the nucellus, and passing
along one of the numerous passageways of the beak meets the
approaching pollen tube (Fig. 90). It is needless to comment

upon this peculiar structure and be-
havior until the details are more fully
known.

Gnetum. — In Strasburger's ac-
count 3 of G. Gnemon it is not apparent
whether or not the mother cells de-
velop a row of potential megaspores.
In any event, several megaspores begin
to enlarge, and free nuclear division
may occur in a number of them before
one finally dominates the rest.

The germination of the megaspore
has been given in detail by Lotsy,15 and
his account for Gnetum Gnemon is as
follows: The nucleus of the megaspore
begins the usual series of free nuclear
divisions, but since there is such great
diversity between the micropylar and
antipodal contents of the embryo sac,
Lotsy raises the question whether the
polarity of the sac may not have been
established by the first division of the
megaspore nucleus. Unfortunately,
the persistent absence of spindles in
his material prevented him from an-
swering the question. Soon the free
nuclei become imbedded in a thin parietal layer of cytoplasm
surrounding a great central vacuole (Fig. 91, IT). In this con-
dition a constriction develops in the sac somewhat below the mid-
dle of its long axis, which persists until the final growth of the sac
obliterates it. This constriction divides the sac into two very dis-
tinct chambers. The smaller antipodal chamber speedily fills
with a compact tissue, in the way usual for the whole embryo

FIG. 90. — Tumboa: the figure
to the right represents a
group of archegonium ini-
tials sending out tubular
prolongations into the nu-
cellus to meet the descend-
ing pollen tubes ; the figure
to the left and above rep-
resents the suspensorlike
elongation of the oospore,
bearing at its tip embry-
onal cells. — After STRAS-

BURGEB.

GNETALES

127

sac in other Gymnosperms, while the micropylar chamber still
contains only free nuclei (Fig. 91, 7).

It is in this stage that fertilization occurs, for the free
nuclei, or presumably free cells, of the micropylar chamber are
all apparently potential eggs. Archegonia, therefore, are no
more developed than in Angiosperms. If fertilization does not
take place, the whole sac becomes filled with tissue. After fer-
tilization the tissue of the antipodal chamber renews its growth,
finally destroys the nucellar tissue, and encroaches upon the rest
of the sac, a small cavity at the micropylar end representing the
micropylar chamber. Lotsy also states that certain of the free
nuclei which are not fertilized organize definite cells, resembling
free endosperm cells of Angiosperms. The same writer also
calls attention to the fact that where the tissue of the antipodal
chamber projects in conical fashion into the constriction certain
structures resembling rudimentary archegonia appear.

The species of Gnetum investigated by Karsten 10' llj 12
show still further reduction of the female gametophyte than that
exhibited by G. Gnemon. In these the whole sac behaves as
does the micropylar chamber in G. Gnemon, so that endosperm
tissue is not developed until after fertilization. It would seem,
therefore, that G. Gnemon represents a condition of the gameto-
phyte intermediate between its structure in Tumboa and the
species of Gnetum investigated by Karsten.

The strong polarity of the sac, the occurrence of naked eggs,
the large or entire development of endosperm tissue after fer-
tilization, are all features suggestive of the Angiosperm condi-
tion. Especially striking in the comparison with the Angio-
sperm embryo sac is the occurrence of free cells at the micro-
pylar end of the sac, and of a compact tissue at the antipodal end.

THE MALE GAMETOPHYTE

The development of the male gametophyte in Gnetales is
practically unknown. In Strasburger's account 3 of Ephedra
campylopoda it is stated that in the germination of the micro-
spore three small cells are successively cut off, the first disor-
ganizing, the second persisting, and the third presumably repre-
senting the primary spermatogenous or generative cell. This is
an ordinary Gymnosperm outline, and the persistence of the
second vegetative cell recalls Giiikgo.

128

MORPHOLOGY OF SPERMATOPHYTES

FIG. 91. — (fnetum Gnemon: A, ovulate strobilus; J5, strobilus upon which only one
seed has matured ; <?, longitudinal section of a young ovulate strobilus, x 24 ; Z>,
later stage, showing ovules and paraphysial hairs (A), x 45 ; E, longitudinal section
of an immature ovulate flower, showing u perianth " ( JP), inner integument (*), rudi-
mentary outer integument (r), and nucellus (ri) ; F, nucellus showing three mother

GXETALES 129

In Gnetum our only knowledge of the male gametophyte
comes from the work of Lotsy 15 upon G. Gnemon, in which the
pollen tube just before fertilization was observed to contain the
tube nucleus and two male cells of the ordinary Conifer type.
Concerning the male gametophyte of Tumboa the only available
information is contained in a few figures by Strasburger.9 The
pollen grain is rather narrowly oblong in section, and on one of
the broad sides of the section a small lenticular cell is shown,
undoubtedly representing a vegetative cell. Whether there is
only a single persistent cell of this kind, as in Cycads, or whether
it is but one of two cells, as in Ginkgo and Conifers, is not appar-
ent. The figures also show the division of the nucleus of the
antheridium initial into tube nucleus and generative nucleus.
Such evidence as is available indicates that the male gameto-
phyte conforms to the type observed among other living groups
of Gymnosperms.

FERTILIZATION

We have no information concerning the fertilization of
E plied ra. The anomalous behavior of Tumboa has already been
referred to, in which the wall of the archegonial initial develops
a tubular process that penetrates the nucellus, meets a de-
scending pollen tube, and receives its discharge. The advance
of archegonial tubes and pollen tubes is expedited by the numer-
ous passageways which riddle the caplike apex of the nucellus.
The egg remains in the bulbous basal portion of the archegonial
initial, and is fertilized there.

The fertilization of Gnetum Gnemon is given in consider-
able detail by Lotsy.15 One or more pollen tubes, each con-
taining a tube nucleus and two male cells, penetrate into the
micropylar chamber of the embryo sac, either at its apex or at
the side. The tip of the tube within the sac usually becomes
much distended, and discharges both male cells through a ter-

cells: G. longitudinal section of a mature ovulate flower, showing the "perianth " (/>),
outer integument (o), and inner integument (/), x 33 ; H, embryo sac with free nuclei in
a parietal layer, and showing the beginning of the constriction which results in the
antipodal and micropylar chambers; /, cell formation in the antipodal chamber, another
sac which is being pushed aside is shown at the right above, x 50 ; •/, upper portion of
sac soon after fertilization, showing a pollen tube with its nucleus (t\ oospores (/), and
unfertilized eggs, x 340 ; K, embryo sac showing the suspensorlike development of the
oospore (a^ advancing toward the antipodal endosperm, x 50. — All from LOTSY except-
ing F, which is after STRASBL-RGER.

130 MORPHOLOGY OF SPERMATOPHYTES

minal pore. Each male cell fuses with a free egg, and as many
oospores are formed as there are male cells discharged into the
sac — that is, twice as many as there are discharging pollen
tubes (Fig. 91, /). The oospore was observed to organize at
once about the fusion nucleus a comparatively dense layer of
cytoplasm, and soon a cell membrane appeared. The oospores
may remain free, or they may become attached to the tips of the
pollen tubes which belong to them.

IV. THE EMBRYO

Ephedra. — An account of the development of the embryo of
Ephedra aUissima. is given by Strasburger.3 Germination be-
gins with free nuclear division, and two to eight free cells, some-
times more, are organized within the oospore (Fig. 89, B).
These free cells do not organize any definite tissue as in Coni-
fers. As a consequence, each cell continues to act independ-
ently, elongating to form a suspensorlike tube, which emerges
from the oospore and penetrates the endosperm (Fig. 89, C).
The more or less numerous tubes, emerging in various direc-
tions, are bulbous within the oospore, where they are in contact
with its nutritive supply, and at the free tip cut off a single cell,
from which the embryo proper is developed. The embryonal
cell divides transversely, which is followed by other transverse
divisions, forming a short filament (Fig. 89, D, E}. Longitudinal
divisions succeed, and later periclinal divisions (Fig. 89, F).- No
details of the organization of the growing points are recorded.

Tumboa. — After fertilization, it is reported that the arche-
gonium initial elongates and penetrates the endosperm as a
suspensor, which is a long, much-coiled, and very persistent
structure, and carries the egg in its tip. It is probably the fact
that the so-called archegonial suspensor is a tube developed by
the oospore, just as in the free oospores of Gnetum, and that the
so-called egg in its tip is the nucleus of the oospore. In any
event, the embryonal cell appears as a small cell cut off at the
tip of the suspensorlike tube (Fig. 90), and gives rise not only to
the embryo proper but also to numerous embryonal tubes
(Fig. 92). ^

According to Bower's account 6j 7 of the embryo, the first
division of the embryonal cell is longitudinal. Then a trans-

GXETALES

131

verse division and other longitudinal divisions follow, until a
pyramidal mass of cells is organized at the end of the suspensor.
From the basal cells of this mass the embryonal tubes develop
(Fig. 92). No details of the organization of the various growing
points are given, but the completed embryo consists of a well-
developed root and a long hypocotyl, between
which there is organized a very conspicuous
footlike process, that remains in the seed dur-
ing its germination. Two cotyledons are de-
veloped, as in all Gnetales.

Gnetum. — According to Lotsy,15 the free
oospores develop long tubes, which advance
toward the mass of antipodal endosperm and
penetrate into it deeply, sometimes branch-
ing. During this advance the nucleus keeps
in the tip of the tube (Fig. 91, A"). There is
every reason to suppose that this is a repeti-
tion of the behavior noted in Tumboa. It
is after this penetration of the antipodal re-
gion that the endosperm renews its growth
and finally encroaches upon the whole nucel-
lus. It is in this condition, with the tubular
oospore, containing the nucleus in its tip, im-
bedded in the antipodal endosperm, that the
seed drops. Lotsy reports that occasionally
the tubes are not directed toward the antip-
odal region, but may pass out of the micro-
pylar chamber into the nucellus, or may grow
directly away from the antipodal region. In
either of these cases there is no further de-
velopment.

The further development of the embryo
of Gnetum Gnemon has been studied by
Bower.8 At the tip of the oosporic tube an embryonal cell is
cut off. This cell divides longitudinally and transversely, until
a small group is organized (Fig. 93, A), the basal cells of
which give rise to embryonal tubes, just as in Tumboa. The
mature embryo also resembles that of Tumboa in its two coty-
ledons, and prominent hypocotyl and root between which a con-
spicuous foot is developed (Fig. 93, B\

FIG. 92.— Tumboa \ a
young embryo con-
sisting of a pyram-
idal group of cells,
the basal ones giv-
ing rise to embry-
onal tubes. — After
STRASBURGEB.

132

MORPHOLOGY OF SPERMATOPHYTES

A

FIG. 93. — G-netum Gnemon:
.4, a young embryo at the
tip of a suspensor and im-
bedded in endosperm ; B,
a germinating seed, the
root and stem tips having
escaped, and the foot re-
maining within the seed
coat. — After BOWER.

In the germination of the oospore, and in the structure of
the female gametophyte, there is evidence that Tumboa and
Gnetum are allied forms, the latter hav-
ing proceeded further in the reduction
of its gametophyte preceding fertiliza-
tion. It is also evident that Ephedra
must be associated with them as the least
modified form, its gametophyte retain-
ing the character of those in other Gym-
nosperms, while the germination of its
oospore is fairly transitional between
Conifers on the one hand and Tumboa
and Gnetum on the other. There is free
nuclear division in the oospore, as in
Conifers, but the cells organized in con-
nection with them remain independent
and behave as do the oospores of Tumboa
and Gnetum. It should be remarked,
also, that the number of free cells de-
veloped within the oospore of Ephe-
dra may be reduced to two, and that any further reduction
would probably result in causing the whole oospore to behave as
do these of Tumboa and Gnetum. If Ephedra be associated
with Tumboa and Gnetum, as seems reasonable, there is equally
weighty evidence for regarding it as the group of Gnetales most
nearly related to the Conifers, for its female gametophyte and
the early stages in the germination of its oospore are identical
with those in Conifers.

LITERATUEE CITED

1. HOOKER, J. D. On Welwitschia, a new genus of Gnetaceae.

Trans. Linn. Soc. London 24: 1-48. pis. 1-14. 1863.

2. EICHLER, A. W. Ueber Welwitschia mirabilis, etc. Flora 47:

459-464, 473-479, 489-496, 508-510, 513-520. 1863.

3. STRASBURGER, E. Die Coniferen und die Gnetaceen. 1872.

4. BECCARI, O. Delia organogenia del fieri feminei del Gnetum

Gnemon L. Nuovo Giorn. Bot. Ital. 7: 91-99. 1877.

5. STRASBURGER, E. Die Angiospermen und die Gymnospermen.

1879.

6. BOWER, F. O. On the Germination and Histology of the Seed-

lings of Welwitschia mirabilis. Quart. Jour. Micr. Sci. 21 :
15-30. pis. 3-4. 1881.

GXETALES 133

7. BOWER, F. O. On the Further Development of Welwitschia mira-

bilis. Quart. Jour. Micr. Sci. 21 : 571-594. pis. 22-23. 1881.

8. BOWER, F. O. The Germination and Embryology of Gnetum

Gnemon. Quart. Jour. Micr. Sci. 22: 278-298. pi. 25. 1882.

9. STRASBURGER, E. Ueber das Verhalten des Pollens und die

Befruchtungsvorgange bei den Gymnospermen. Hist. Beitr. 4 :
1-158. pis. 1-3. 1892.

10. KARSTEN, H. Beitrage zur Entwickelungsgeschichte der Gattung

Giietum. Bot. Zeit. 50: 205-215, 221-231, 237-246. pis. 5-6. 1892.

11. KARSTEN, H. Untersuchungen iiber die Gattung Gnetum. Ann.

Jard. Bot. Buitenzorg 11 : 195-218. pis. 17-19. 1893.

12. KARSTEN, H. Zur Entwickelungsgeschichte der Gattung Gnetum.

Conn's Beitr. z. Biol. d. Pflzn. 6: 337-382. pis. 8-11. 1893.

13. JACCARD, P. Le developpement du pollen de VEphedra Helvetica.

Archives des sciences physiques et naturelles. III. 30: 280-282.
1893. See review in Bot. Centralbl. Beihefte 4: 230. 1894.

14. JACCARD, P. Recherches embryologiques sur TEphedra Helvetica

(Inaug. dissertation. Zurich 1893). Bull. Soc. Vaudoise des sci.
nat. 30 : pis. 10. March, 1894. See review in Bot. Centralbl. 61 :
111-113. 1895.

15. LOTSY, J. P. Contributions to the Life-history of the Genus Gne-

tum. Ann. Jard. Bot. Buitenzorg II. 1 : 46-114. pis. 2-11. 1899.

CHAPTER V

FOSSIL GYMNOSPEBMS

THE numerous fossil remains, from the early Paleozoic
onward, which have been referred to Gymnosperms, have been
determined chiefly by their anatomical features. This result
has come from the nature of the available material, but to the
morphologist it is suggestive rather than demonstrative. A vast
amount of material has been named, which has been made some-
what available for the morphologist by Solms-Laubach,3 and
most recently by Scott.6 Unfortunately, at this writing the
second volume of Seward's Fossil Plants, which will deal with
Gymnosperms, has not appeared. In any consideration of the
phylogeny of Gymnosperms the historical point of view is abso-
lutely essential, and in so far as there are any real morphological
data among paleobotanical material they must be considered in
the present connection.

Paleobotanists have recognized a group of fossil forms inter-
mediate between Ferns and Cycads, to which Potonie has given
the name Cycadofilices. They were especially abundant during
the Carboniferous, and many of them have been described as
Ferns. Some of the characteristic forms included in this pro-
posed intermediate group are Heterangium, Lyginodendron,
and Medullosa. The testimony in reference to the position of
these forms is entirely anatomical, but it is extremely suggestive
of the intermediate position claimed for the group by most paleo-
botanists. The historic evidence seems to be clear that the
paleozoic Ferns merged gradually into the Cycads.

In addition to these transition forms, concerning which the
morphologist must hold his judgment in suspense until the more
essential structures have been discovered, there are certain ex-
tinct groups whose Gymnosperm character is undoubted.
Among them the most conspicuous are the Cordaitales.
134

FOSSIL GYMNOSPERMS

135

CORDAITALES

The amount of information that has been accumulated con-
cerning this great Paleozoic group is a matter of surprise. Not
only are the vegetative structures abundantly preserved, even in

FIG. 94. — Cordaites laevis: branch (restored) bearing leaves and strobili, and also a
large bud. — After GRAND' EURY.

their minutest details, but the more telling reproductive struc-
tures are known sufficiently to indicate general relationships.
The Cordaitales become evident in the Silurian, and increase
10

136 MORPHOLOGY OF SPERMATOPHYTBS

in display to the Carboniferous, when they were astonishingly
abundant, the leaves often being packed together in masses on
every layer of the deposits. It seems clear that this was the pre-
vailing Gymnosperm forest type of the Paleozoic. They were
tall, rather slender trees, with shafts from 10 to 30 meters in
height, branching only above, and with a dense crown of
branches on which simple large leaves were produced in great
abundance (Fig. 94). The leaves were usually long, ribbonlike,
and parallel-veined, from 20 to 100 centimeters long and 15 to
20 centimeters broad. Scott 6 says that " the habit of the Cordai-
tales must have been different from that of any trees with which
we are now familiar. The species with comparatively short
leaves may be compared with such Coniferae as Dammara (the
Kauri Pine of New Zealand), or with some forms of Podo-
carpus, and these trees may best serve to give us some idea of
the extinct family. But the large-leaved species must have had
a habit very different from anything which we are accustomed
to associate with the Gymnosperms at the present day."

The anatomical character of the stem is in general plan that
of the Conifers, but the very large pith, which appears in trans-
verse diaphragms, is more suggestive of the Cycads. As in Coni-
fers, the tracheary tissue begins with spiral vessels lying next
to the pith, and the secondary wood is composed exclusively
of tracheids with bordered pits which are usually in two or more
rows and are densely crowded, so that they have a hexagonal out-
line. As this secondary wood resembles very closely that of
Araucaria, it was named Araucarioxylon by Kraus. The cor-
tex contains strands of fibrous sclerenchyma, and isolated gum
receptacles.

The leaves have conspicuous parallel veins, and in the major-
ity of forms are elongated and strap-shaped, having very much
the appearance of such monocotyledonous leaves as those of
Yucca or Dracaena. The venation is repeatedly dichotomous,
excepting in the narrowest leaves, which are almost grasslike.
Sometimes, however, the leaves are short and obovate, and are
said occasionally to branch dichotomously and even to become in-
cised. In such leaf forms and venation there are suggestions of
Ginkgo and of certain Ferns, while the ordinary elongated type,
with its parallel veins, suggests Tumboa.

The anatomical character of the leaf is practically that of

FOSSIL GYMXOSPERMS

Cycads, and Scott calls attention to the fact that the structure of
the leaf of Cordaites resembles that of a single pinna of the leaf
of such a Cycacl as Zamia, or one of the larger-leaved Arau-
carias. Leaf sections show a very xerophytic structure, hypo-
dermal masses of s.clerenchyma producing great firmness, if not
rigidity. In the interior of the leaf, between the bundles,
ki transfusion tissue " is developed, the cells being elongated par-
allel to the epidermal layers, and so loosely organized as to
appear like thin diaphragms between very large intercellular
spaces, just as in the Cycads and Sciadopitys. Each vascular
bundle is surrounded by a strong sheath, which connects with the
hypodermal mass of sclerenchyma. Each bundle is of the
mesarch type — that is, the secondary xylem occurs in two parts,
with the primary spiral elements between them. This is exactly
the structure of the leaf bundles of Cycads, and of other Paleo-
zoic plants. The numerous stomata are depressed and occur in
rows, apparently only upon the under surface. It is an interest-
ing fact that the form and general structure of the leaf suggest
the leaves of Conifers, while the anatomical features are more
those of the leaves of Cycads.

The roots which are regarded as belonging to the Cordaitales
show structures which relate them easily to the stem. In the
various specimens recorded the vascular cylinder of the root is
cliarch, triarch, or tetrarch.

Chiefly through the researches of Renault, some knowledge
of the strobili of a few species has been obtained. The strobili
occur upon special shoots, which may be small or branched, and
bear small, catkinlike clusters (Fig. 94), which do not often
exceed a centimeter in length. Externally the staminate and
ovulate clusters are alike, although in some cases an ovulate
strobilus may consist of a solitary ovule with few bracts. Each
strobilus usually consists of a series of prominent leaflike bracts,
in the axils of which the spore-bearing structures occur. It
should be borne in mind that the following account is derived
from the examination of a few sections of one or two species of a
very large and probably diverse group.

The staminate strobilus consists of a thick axis covered by
long spirally arranged bracts, which resemble reduced leaves.
Many of the bracts are sterile ; but especially toward the apex
of the strobilus these are replaced by structures which may be

138

MORPHOLOGY OF SPERMATOPHYTES

fairly regarded as sporophylls. These sporophylls are narrower
than the sterile bracts, and each bears at the apex a cluster of two
to six sporangia, which are linear and erect, and dehisce by a lon-
gitudinal slit (Fig. 95, B). The sporophyll character of this
structure is the view suggested by Renault, and according to this
view it is morphologically a stamen. Solms-Laubach, however,
regards each of these so-called stamens as a staminate flower, the

\ X "V ~^' ^> ^f

^^5&^^^SU^//

FIG. 95. — Cordaites, staminate strobilus: A, transverse section; ^longitudinal section
of strobilus of C. Penjoni ; (7, a pollen grain, showing the group of internal cells.—
After RENAULT.

sporangia being sessile upon the summit of a pedicel. Such a
view would suggest, as a parallel, the staminate flowers of Eplie-
dra and Gnetum, in which two or more sessile sporangia are borne
by a special axis which may be regarded as a pedicel, or at least
a shoot. Scott suggests, however, that the whole cluster of
strobili corresponds more to the staminate cluster of Giiikgo
than to any other known Gymnosperm. In any event, the in-
florescence is different from that of any other Gymnosperm, and

FOSSIL GYMNOSPERMS 139

suggests a condition which needs further material before it can
be homologized with any of them.

The pollen grains are remarkably well preserved, the exine
being cuticularized and covered with a fine " shagreenlike "
reticulation. As these spores have been found both in the spo-
rangia and in the pollen chamber, there can be no doubt as to
their relation to the ovules examined. The most remarkable
feature of the group is the presence of a distinct tissue within
the pollen grain while still in the sporangium. This tissue con-
sists of a group of small polygonal cells against the wall on one
side (Fig. 95, C). After pollination, while in the pollen cham-
ber, the spore is said to increase very much in size, and the inter-
nal tissue develops still further. If these statements can be re-
lied upon, the male gametophyte of Cordaitales is one of extraor-
dinary interest, for it is far more primitive in character than that
of any other known Gymnosperm, and points to the earliest stages
of heterospory, earlier than in any existing heterosporous Pteri-
dophyte. Whether the cells of this internal tissue are vegetative
or sperm-producing is a matter of conjecture. In any event,
the reduction of the male gametophyte is very much less than in
any other known Seed-plant. Taken in connection with the pres-
ence of ciliated male cells in the Cycads and Ginkgo, and of a
very deep pollen chamber which eventually deepens down to the
embryo sac, it might be inferred that in Cordaitales there was
little if any development of a pollen tube, and that swimming
male cells, if not actually sperms, were discharged within easy
reach of the archegonia. It must be remembered that in
the Cycads and Ginkgo the pollen tube does not serve the pur-
pose of a sperm carrier, and there is no reason to suppose that
there was any such method of sperm transfer in the Cor-
daitales. In case the internal tissue of the pollen grain is
antheridial, it would indicate the production of more numer-
ous male cells than in any other Seed-plant. The whole struc-
ture is so suggestive of a theoretical transition stage between
Pteridophytes and Spermatophytes that it is a temptation to
make more out of it than the amount of observed material may
justify.

The ovules occur in strobili similar to those which bear the
stamens, but the bracts are longer and more overlapping, and a
terminal tuft of long bracts crowns the strobilus. The solitary

140

MORPHOLOGY OF SPERMATOPIIYTES

ovules are evidently cauline, being borne at the summit of a short
bractlet-bearing shoot, which arises from the axil of a bract. As
has been mentioned before, the ovules may be reduced to a soli-
tary one in a strobilus, and the bracts may be very few in num-
ber. Two envelops are ap-
parent, and permit the
same diversity of opinion as
exists in reference to the
ovular envelops in the Gne-
tales. It seems altogether
simpler to regard the two
envelops in Cordaites as
two integuments, although
there is no tubular pro-
longation of the inner in-
tegument as in the Gne-
tales. The nucellus devel-
ops a remarkably promi-
nent and persistent beak,
composed of modified and
heavy-walled cells, through
which a passageway leads
to a large pollen chamber
(Fig. 96). The whole
structure of the nucellus
recalls the Cycads and

FIG. 96. — CordiantJtus Grand"1 Euryi, showing
the beak of the nucellus ; wedged in the pas-
sagewa3T are the large pollen grains, which
show the shagreenlike surface and the in-
ternal group of cells. — After RENAULT.

Ginkgo, and, as is the case
with them, the nucellar tissue which caps the embryo sac
disorganizes, permitting the persistent beak to settle down
upon the embryo sac (Fig. 97). There is also a striking
resemblance to the nucellus of Ginkgo, as described by Hirase,
in the fact that the endosperm develops a protuberance which
supports the settling beak like a " tent-pole." Under the shel-
ter of this improvised tent the archegonia are found and the
male cells are discharged. The situation is so suggestive of
Cycads and of Ginkgo that one can hardly escape the conviction
that in the process of fertilization they are approximately the
same. Traces of the archegonia have been found, so that their
position is definitely determined.

The ovulate structures of the Cordaitales as described above

FOSSIL GYMNOSPERMS

141

are very suggestive of Taxus. The terminal ovule upon a short
bractlet-bearing axis seems to be common to both. The simi-
larity is increased by the fact that in most of the Cordaitales
the testa develops into a fleshy outer layer and a stony inner
one, which suggests the fleshy aril and stony seed coat of Taxus.
Abundant remains of seeds are found, which apparently belong
to the Cordaitales. In most cases the testa develops as indi-
cated above, and closely resembles the seeds of Cycads and
Ginkgo, as well as certain Conifers. In other cases, however,
the seeds ripen dry and bear wings, which are exceedingly vari-
able in form and position. Curiously enough, while seed sec-
tions have brought to light most of the structures referred to
above, no embryos have been found. Probably their for-
mation was much delayed, as in Ginkgo, Gnetum, and some
Cycads.

It seems evident that we are dealing here with a group of
Gymnosperms which deserves to be set apart as of equal rank

mt

ap

ap

FIG. 97. — Cycadinocarpus Augustodunensis, showing upper part of ovule in longitudinal
section : int, integument ; mi, micropyle ; pc, pollen chamber containing pollen
grains ; nu, remains of nucellus ; ar, archegonia ; pr, endosperm. — After RENAULT.

with Cycadales, Ginkgoales, and Coniferales. It certainly
presents important characters which serve to define it as clearly
as any of the living groups, and its ancient character suggests
that it must have been very intimately associated with the begin-
nings of the Gymnosperm phylum.

14:2 MORPHOLOGY OF SPERMATOPHYTES

BENNETTITALES

Cordaitales were the dominant Paleozoic Gymnosperms,.
and associated with them were the numerous Cycadofilices,
whose synthetic character is indicated by the name. The true
Cycadean type, however, does not become apparent until the
Mesozoic; and all through that age, from the Triassic to the
Lower Cretaceous, there was an extraordinary display of Cyca-
dean forms. In anatomy and general vegetative characters the
resemblance to living Cycads is undoubted, but in the ma-
jority of cases the spore-producing structures are remarkably
different. These so-called anomalous Cycadean forms of the
Mesozoic have been taken to represent a fifth great Gymno-
sperm phylum, which constituted the dominant vegetation of
the Mesozoic, much as the Dicotyledons dominate the flora
of to-day.

The remains of Bennettitales have been found in abundance
in the explored regions of both hemispheres, and they have been
described under a variety of generic names, Bennettites, Cycade-
oidea, and Cycadella being the principal ones. The trunk is of
the tuberous type, and is covered by an armor of closely imbri-
cated persistent leaf bases. This character, together with the
histological details of the stem tissues, the arrangement, form,
and anatomy of the leaves, all have their counterpart in living
Cycads. In two important respects, however, the Bennettitales
differ from the Cycadales in external features. One is in the
occurrence of numerous short, lateral, and probably axillary
shoots which bore the inflorescence ; the other is in the presence
of scalelike hairs, which were packed between the leaf bases
and about the inflorescence and its bracts, and resemble the so-
called " ramenta " of Ferns rather than any known structure in
Cycadales. The course of the leaf traces in Bennettitales differs
decidedly from that in living Cycadales. In the latter, two
bundles leave the stele near together, " and curving in opposite
directions pass nearly halfway round the stem, thus entering
the leaf base on the opposite side from their starting points."
Scott 6 describes the leaf traces in Bennettitales as follows: " A
single bundle leaves the ring, starting from the lower angle of
one of the meshes. ... As the leaf trace passes out through the
cortex, it assumes a horseshoe form, with the concave side in-

FOSSIL GYMNOSPERMS

143

ward. It then breaks up into about twenty smaller bundles,
which enter the base of the leaf."

The ovulate inflorescence, however, is most remarkable, and
unlike anything known in Gymnosperms. The description
of Bennettites Gibsonianus given by Carruthers l and by Solms-
Laubach,2 and found to apply to other forms, is as follows: The
strobilus is a pear-shaped body, which terminates a short axil-
lary shoot from the main stem,
and in the species cited is about
5 centimeters long. Only fully
ripened specimens, as shown by
the seeds, have been examined.
The strobilus is completely in-
vested by bracts, which spring
from the shoot and close in over
the apex (Fig. 98).

An account of the details of
the structure is adapted from
Scott 6 as follows : The stalk is
expanded into a hemispherical
receptacle, on which all the or-
gans of the fruit are inserted.
From the convex surface of the
receptacle spring a great number
of slender stalks, which pass ver-
tically upward, or diverge slight-
ly toward the curved surface of
the fruit. Each of these stalks
bears at its end a single erect
seed, with the micropyle directed
outward. The seeds are so placed

FIG. 98. — Bennettites Gibsonianvs, dia-
gram of ovulate strobilus : re, re-
ceptacle ; br, ensheathing sterile
bracts ; s, seeds, each borne on a
long stalk which arises from the
receptacle, and in each an embryo
with two cotyledons ; p, dilated
ends of the interseminal scales,
which become confluent. — Modi-
fied by SCOTT, after SOLMS-LAU-
BACH and POTONIE.

that the longitudinal axis always

meets the surface of the fruit approximately at a right angle.
The spaces between the stalks are packed with scales, which are
dilated at their distal ends, between the seeds, so as to form
a continuous envelop, only interrupted by narrow pits, into which
the seeds exactly fit. The surface of the strobilus, therefore,
appears as a continuous tissue, which is perforated only by the
micropyles of the seeds.

The whole structure presents the appearance of a remarkably

144

MORPHOLOGY OF SPERMATOPHYTES

d'.

modified strobilus, whose axis has become shortened and dilated
into a receptacle, resulting in a massing of bracts and axillary
sporangiferous shoots, the bracts in ripening organizing by

their tips a continuous investment.
Whether this peculiar character of
the bracts represents a post-fertiliza-
tion change, or is developed before
pollination, is a question for future
discovery to answer. In any event,
the solitary cauline ovule, terminal
upon an axis of variable order, seems
to represent an ancient Gymnosperm
condition.

From the testa of the seed it is
impossible to determine the nature of
the integuments. The testa is closely
adherent on the outside to the inclos-
ing and confluent tips of the bracts,
and shows three layers, the middle
one composed of very thick-walled
tissue, and much dilated toward the
micropyle. The innermost layer
forms the tubular micropyle, which
becomes somewhat dilated at the apex
and opens upon the surface of the
strobilus. It would appear that two
integuments are indicated, an inner
delicate one continued into the tubu-
lar micropyle, and a much heavier
outer one. The third or outermost
layer of the testa, which is closely adherent to the surrounding
bract tissue, may represent either an external layer of the outer
integument, or a limiting layer of the bract tissue. Such an in-
terpretation would conform in general with the structure of the
ovule in all the more primitive Gymnosperm types.

Longitudinal sections of the seeds have also revealed a nu-
cellar beak and a large pollen chamber, which suggests Cordai-
tales, Cycadales, and Ginkgoales. One of the most striking
characters revealed by the seed, however, is the occurrence of a
large dicotyledonous embryo completely filling the embryo sac,

FIG. 99. — Bennettites Gribsonia-
nus, longitudinal section of
seed : a, the entering vascular
bundle; 6, its expansion at
base of nucellus ; c, the em-
bryo with two unequal coty-
ledons ; rf, the testa, terminat-
ing above in the tubular pro-
cess (r/), which leads to the
surface ; e, probably remains
of the nucellar beak, contain-
ing the pollen chamber. —
After SOLMS-LAUBACH.

FOSSIL GYMNOSPERMS

145

in fact the whole seed with the exception of the beak (Fig. 99).
The evident absence of a suspensor at once suggests Ginkgo, but
the nearly complete disappearance of the endosperm is without a
parallel among Gymnosperms, so far as studied.

The morphological features of Bennettitales have been ob-
tained heretofore almost exclusively from European material,
but the remarkably rich display of this group in the Mesozoic
flora of the United States, as developed by Professor Lester F.
Ward and his associates, promises to supply much additional in-
formation. Large collections of this material are in the posses-
sion of the L'nited States National Museum, and of the Yale
University Museum, and the minuter structural characters are
being investigated by Dr. G. R. Wieland, who has published pre-
liminary papers concerning the " male flower " 4 and the " fe-
male fructification " 5 of species of Cycadeoidea. In some of
the species, as C. Wielandi, the ovulate strobilus seems to con-
form almost exactly to the description of the strobilus of Bennet-

FIG. 100. — Cycadeoidea Wielandi, transverse section of a portion of the ovulate strobilus,
showing the seed-stalks with their central bundles, and the interseminal scales
packed in between. — From a photograph by C. E. BEECHER from a section made by
G. K. WIELAND from a specimen in Yale University Museum.

tites Gibsonianus, as given above. There is the same broad
receptacle bearing thickly set stalked seeds and interseminal
scales (Fig. 100), and all ensheathed by overlapping bracts to
form an ovoid body (Fig. 98).

The discovery of microsporophylls and of bisporangiate

146

MORPHOLOGY OF SPERMATOPHYTES

strobili,7 however, are the most interesting of Wieland's results
thus far. The staminate strobili originally described4 were
taken from specimens of Cycadeoidea ingens, and apparently
resemble the ovulate strobili of the group in all important fea-
tures. Each strobilus is borne
terminally upon an axillary
dwarf shoot, and is ensheathed
by a cluster of bracts. There
is plain indication of the ex-
istence of a thick central axis
upon which the numerous
sporophylls were crowded,
each sporophyll bearing nu-
merous sori. This first mea-
ger description of a staminate
strobilus has been supple-
mented by lattr preparations
of Wieland from the strobili of
the remarkable Cycadeoidea
Dacotensis.5 "With great gen-
erosity Dr. Wieland has placed
these recent preparations at
our disposal, and from them the
following account is written:

The strobilus of Cycadeoi-
dea Dacotensis is of the usual

ovoid form, being completely ensheathed by overlapping sterile
bracts. Just within the ensheathing bracts there is a set of very
prominent microsporophylls, whose abaxial surfaces are densely
covered by well-defined sori (Fig. 101). As is evident from the
figure, and much more conspicuously so in other preparations,
the sori occur closely packed together in well-defined lines. The
section of a sorus reveals characters unknown heretofore among
Gymnosperms, and remarkably similar to those of Marattiaceae
(Fig. 102). The structure seems to be almost identical with
that of the synangium of Marattia or of Danaea, and presents
more the appearance of a plurilocular sporangium than of a
sorus. Superficially there is a wall of heavy cells, and between
it and the regular sporangial chambers there is a more delicate
tissue. The sporangial chambers, separated from one another

FIG. 101. — Cycadeoidea Dacotensis, a small
part of a radial oblique section of a stro-
bilus, showing some of the numerous
synangia, x 8. — From preparation made
by G. R. WIELAND from specimen in
Yale University Museum.

FOSSIL GYMNOSPERMS

by septa consisting of a single plate of cells, extend the full
length of the broad synangium — that is, from its base to its
apex, and form two rows, one on each face of the synangium,
probably separated from each other by a few layers of inter-
vening tissue. There is every evidence that the whole struc-
ture at Urst consisted of a mass of homogeneous tissue, in which
regularly placed sporogenous masses were differentiated, each
mass standing for a sporangium, exactly as in Marattia and in
Danaea. At maturity the tissue between the two rows of
sporangia apparently breaks down, at
least the synangium splits into two
valve?, just as in the genera of Maratti-
aceae referred to. In Fig. 101 trans-
verse sections of the two rows of sporan-
gia may be seen in most cases, and in
a few instances a different point of view
has made visible the septa which sepa-
rate the sporangial chambers. In Fig.
103 this is especially evident, where a
single synangium presents a somewhat
striking resemblance to the trabeculate
sporangium of Isoetes.

AVe have used the terms sorus and
synangium indiscriminately in describ-
ing these sporangial structures, upon the
hypothesis that a synangium and a sorus
are morphological equivalents. We
recognize the fact that such a synan-
gium as we have described may be re-
garded as a single plurilocular sporan-
gium, resulting from the sterilization of
potential sporogenous tissue, as in Iso-
etes. To discuss this question, however,
does not fall within the purpose of this
book. The important fact to the stu-
dent of Gymnosperms is that among
the Bennettitales there are forms whose microsporangiate struc-
tures are almost identical with those of such Ferns as Marattia
and Danaea, whether these structures be called sori, synangia,
or plurilocular sporangia.

FIG. 102. — Cycadfoidea Daco-
tensis. transverse section of
a small portion of a synan-
gium, showing outer layer
of heavy-walled cells, the
inner more delicate tis-
sue, and three sporangial
chambers containing pollen
grains (p).

148

MORPHOLOGY OF SPERM ATOPHYTES

FIG. 103. — Cycadeoidea Dacotensis,
a very small part of a some-
what tangential longitudinal
section of a strobilus, to show
in side view the septa of a syn-
angium, x 8. — From prepara-
tion made by G. R. WIELAND
from specimen in Yale Univer-
sity Museum.

We have described the strobilus
of Cycadeoidea Dacotensis as consist-
ing of a sheath of overlapping bracts,
within which microsporophylls bear-
ing Marattialike synangia occur.
Within the microsporophylls, and
towards the base of the strobilus,
there is discovered what at first
sight appears to be merely the conic-
al apex of the axis of the strobilus,
but which proves to be the ovulate
region (Fig. 104). The conical re-
ceptacle is covered all over with the
characteristic stalked seeds and " in-
terseminal scales." In this case the
" receptacle " really resembles an
axis, and is of the ordinary Gymno-
sperm type; while in Bennettites

and certain species of Cycadeoidea the receptacle is flattened.

This elongated ovulate axis is not always associated with a bi-

sporangiate strobilus, as it occurs in strictly ovulate strobili.

In Fig. 105 there is

shown a transverse

section of a young

ovulate strobilus of

Cycadeoidea colossal-

is, whose axis is of

the elongated type, as

in the bisporangiate

C. Dacotensis (Fig.

101), but which is

invested directly by

the ensheathing ster-
ile bracts. The illus-
tration also shows

some of the character- ,

, . , FIG. 104. — Cycadeoidea Dacotensis, longitudinal section

StlC ramenta Which through basal portion of strobilus, showing at the

Were packed between left portions of synangium-covered sporophylls, and

the superficial sheath at the right the ovulate region of the strobilus' x 4-

j ,£ ,. /. — From preparation made bv G. R. WIELAND from

and the OVUlllerOUS specimen in Yale University'Museum.

FOSSIL GYMXOSPERMS

149

part of the strobilus. The two distinct ovules show that they
are young by the fact that their stalks have not elongated.

The monoecious habit among Gymnosperms is not unusual,
although it has never been found among Cycads; but that this
habit should take the form of a bisporangiate strobilus is espe-
cially interesting. Such strobili are found occasionally among
Conifers, but they are regarded as abnormal (Fig. 59). To
discuss the significance of these bisporangiate strobili seems to
us premature at present, since the results thus far announced
are very meager as compared
with those we have a right to
expect. It may be well, how-
ever, to make the following
statements. That a bispo-
rangiate strobilus must have
existed at some time among
the ancestral forms of Gym-
nosperms has already been in-
dicated by Tumboa, in which
the association of the two
forms of sporangia is more
intimate than in Cycadeoidea.
It seems evident, also, that a
bisporangiate strobilus holds
no more relation to an angio-
spermous "flower" than does
a monosporangiate strobilus;
and that neither type of stro-
bilus among Gymnosperms is
related to the angiospermous
flower in any way. Whether
a bisporangiate or a monosporangiate strobilus is the more
primitive is a question that probably finds its answer in the
statement that both conditions exist in the most primitive
forms. In Bennettitales, therefore, we are discovering inter-
esting transitions between Pteridophytes and living Gymno-
sperms, but there is not as much suggestion of Angiosperms as
the living Gymnosperms have already given.

While the characters of Bennettitales justify the belief that
they must be closely associated with Cycadales, the axillary

t

FIG. 105. — Cycadeoidea colossalis, transverse
section through the central region of a
very young strobilus, showing the axis-
bearing stalked ovules and interseminal
scales, x 10. — From preparation made by
G. R. WIELAND from specimen in Yale
University Museum.

150 MORPHOLOGY OF SPERMATOPHYTES

dwarf shoots, the synangia, the remarkably modified and com-
plex ovulate strobili, the occasional bisporangiate strobili, and
the entire replacement of endosperm by an embryo which devel-
ops without a suspensor, are characters which strongly confirm
the suggestion that the group should be regarded as one of the
main divisions of the Gymnosperms.

CYCADALES

During the Mesozoic, along with the Bennettitales, the
Cycadales undoubtedly existed, but the evidence is scanty. The
reputation of the Mesozoic as " the age of Cycads " seems to have
come rather from the abundant remains of Bennettitales. Cer-
tain remains may be taken fairly to represent Cycads, and doubt-
less numerous extinct genera of Cycads may have existed, but
no modern genus seems to have an unclouded title in the Meso-
zoic except Cycas, which has been recognized in the Rhaetic
beds between the Lias and the Trias. The antiquity of Cycas
confirms morphological opinion concerning it, as it is well sepa-
rated from the other Cycads in features which are recognized
as primitive.

GINKGOALES

The characteristic foliage of Ginkgo makes it easy of recog-
nition, although some forms which have been referred to it may
prove to be Ferns. However, putting together the evidence of
foliage, flowers, and seeds, it is evident that the group was a very
widespread one in the Jurassic, and that the testimony is very
strong for claiming its existence during the Paleozoic. The in-
definiteness of its occurrence in the Paleozoic is based upon the
fact of its association with the great group Cordaitales, into
which it seems to merge. The conclusion reached by Scott 6
is as follows : " On the whole, the sum of the fossil evidence is
of sufficient weight to prove the great antiquity of the gymno-
spermous family now represented by the Maidenhair tree, which
appears to be best regarded as the one surviving member of an
ancient stock, derived from the same cycle of affinity as the
Paleozoic Cordaiteae, once the dominant type of Gymno-
sperms.77

FOSSIL GYMNOSPERMS 151

CONIFEKALES

It would be a matter of great interest if the relative an-
tiquity of the various groups of Conifers could be determined.
The varied structures of this dominant group are extremely
perplexing to relate phylogenetically, and some definite histori-
cal data would go far toward suggesting the possible sequence.
Upon morphological grounds one might be tempted to regard
the Taxeae as the most primitive of Conifers, and Abieteae, espe-
cially Pinus, as the most highly specialized and recent, but so
far as there is any historical testimony it is diametrically
opposed to such a view.

The remains of undoubted Conifers occur from the later
Paleozoic (Permian) on, and the total amount of material is
enormous. Elaborate studies have sought to determine this
material upon the basis of anatomical structure, and the litera-
ture of paleobotany is full of names suggestive of affinities with
modern genera, but these methods have proved to be abso-
lutely untrustworthy. The sure evidence of well-preserved
strobili associated with vegetative structures is lacking, so that
the ancient history of this great group is still in a chaotic
condition.

That the phylum is a very ancient one is evidenced by the
fact that it is known to have existed during the later periods of
the Paleozoic, and possibly earlier, but to what modern group
these Permian forms are most nearly allied is absolutely un-
known. Their resemblance to the Araucarieae in general vege-
tative habit has suggested an Araucarian affinity, but such evi-
dence as the strobili furnish contradicts this.

Leaving these earliest Permian forms of doubtful affinity,
the first suggestion of a modern group is furnished by the extinct
genus Vottzia of the Upper Permian and Triassic, which has
been referred, apparently on fairly safe evidence, to the Tax-
odieae, approaching Cryptomeria in the lobed character of
the ovuliferous scale. True Araucarieae and Cupresseae can
not be traced with certainty below the Jurassic; Abieteae,
allied to Pinus, are not recognized earlier than the "Weal-
den; while Taxeae have not been discovered lower than the
Cretaceous.
11

152 MORPHOLOGY OF SPERMATOPHYTES

It is hardly possible that such a sequence represents the
relative age of the distinct differentiation of these groups, for
it can not be corroborated upon any theory of morphological
relationship.

LITERATURE CITED

1. CARRUTHERS, W. On gymnospermatous Fruits from the Second-

ary Rocks of Britain. Journal of Botany 5 : 1 seq. 1867.

2. SOLMS-LAUBACH, H. Ueber die Fructification von Bennettites

Gibsonianus Carr. Bot. Zeit. 48 : 789-798, 805-815, 821-833, 843-
847. pis. 9-10. 1890.

3. SOLMS-LAUBACH, H. Fossil Botany. 1891.

4. WIELAND, G. R. A Study of some American Fossil Cycads. Part

I. The Male Flower of Cycadeoidea. Am. Jour. Sci. IV. 7:
223-226. pis. 2-4. 1899.

5. WIELAND, G-. R. A Study of some American Fossil Cycads. Part

III. The Female Fructification of Cycadeoidea. Am. Jour. Sci.

IV. 7: 383-391. pis. 8-10. 1899.

6. SCOTT, D. H. Studies in Fossil Botany. 1900.

7. WIELAND, G. R. The Yale Collection of Fossil Cycads. The Yale

Scientific Monthly 6: March, 1900.

CHAPTER VI

COMPARATIVE MORPHOLOGY OF GYMNOSPERMS

AFTER having considered the details of structure presented
by the living and extinct groups of Gymnosperms, it will be
well to summarize them, and to obtain some conception of Gym-
nosperms as a single well-defined assemblage of forms. In this
way the contrasting characters of the different groups may be
emphasized, and the relations of Gymnosperms to Pteridophytes
on the one hand, and to Angiosperms on the other, may be made
to stand out more sharply.

I. THE VEGETATIVE ORGANS

THE STEM

All known Gymnosperms are woody plants, almost all being
trees. Whether the group ever contained herbaceous members
or not is an interesting but an unanswerable question. The
short tuberous stems of many Cycads, of Benne^titales, and of
Tumboa', the simple, palmlike column of certain Cycads; the
tall trunk of the Cordaitales, branching only above ; the excur-
rent shaft of most of the Conifers and of Ginkgo with its sym-
metrical monopodial branching; the trailing and spreading
habit of certain Conifers; the straggling shrubby habit of
Ephedra] and the high climbing habit of certain species of
Gnetum, is a list which seems to exhaust the possible habits of
woody stems. Amidst all this great variety of form there is a
uniform xerophytic structure, indicating adaptation to rela-
tively hard conditions of living. The habitats of Gymnosperms
to-day indicate that they either are not at home in the more
genial conditions affected by Angiosperms, or have not been
able to maintain themselves in competition with this great
group. In any event, it seems clear that the conditions for plant

153

154: MORPHOLOGY OF SPERMATOPHYTES

life are more favorable now than at any previous time in the
history of plants.

The anatomy of the stem shows characters in common with
Pteridophytes, others suggestive of Dicotyledons, and still others
peculiar to the group. In all Gymnosperms the primary vascu-
lar bundles are open collateral, contain true tracheae, and are
organized into a hollow cylinder inclosing a more or less exten-
sive pith region. At this point the resemblance to the dicoty-
ledonous stem ceases. In many Cycads the primary cam-
bium is very short-lived, a succession of secondary cortical cam-
bium cylinders developing successive sets of cauline bundles.
This development of successive cortical cylinders of vascular
strands seems to represent a very ancient type of stem. It is
displayed by Cycas, Encephalartos, Macrozamia, and Bowenia
among the Cycads, belonged to the Bennettitales, and is appar-
ently shared by Gnetum and Tumboa. Furthermore, these sec-
ondary vascular bundles are of the concentric type in Cycas,
and give evidence in Macrozamia and Bowenia of a transition
from the concentric to the collateral type. The concentric type
of bundle is also displayed in the peduncles and leaves of other
genera, as Stangeria, Zamia, and Ceratozamia, as well as in
Bennettitales. It is significant, also, that this type of stem
characterized the so-called Cycadofilices.

The gradual abandonment of the cortical cambiums, and the
conversion of concentric into collateral bundles, seems to have
taken place in the Cycads, as evidenced by Zamia, Dioon, and
Stangeria. In Cordaitales, Coniferales, and Ephedra, the pri-
mary cambium develops all of the secondary vascular tissue, and
the collateral type of bundle is fixed. It is interesting to note
that Cordaitales, the most ancient distinct group of Gymno-
sperms, are included in this last list, and it indicates an improb-
ability that there is any definite and single type of Gymnosperm
stem that can be regarded as the most ancient.

The development of tracheids with bordered pits has long
been regarded as a distinctly Gymnosperm character, but they
seem to be wanting in Tumboa, and are said to be present in
certain fossil Pteridophytes. In any event, it is a type of vessel
displayed almost exclusively by Gymnosperms, and almost with-
out exception among them. A transition from tracheids with
bordered pits to true tracheary vessels seems to be evident in

COMPARATIVE MORPHOLOGY OF GYMNOSPERMS 155

Ephedra, if not in the other Gnetales, while the very regular
massing of these tracheids to form the secondary wood is charac-
teristic of Conifers.

It is also worthy of note that there are no distinct nodes in
the stems of Cycadales and Bennettitales ; that they are usually
evident, but often indistinct in Ginkgoales, Cordaitales, and
Conif erales ; and that in Gnetales they are distinctly developed.

THE LEAF

The spiral and cyclic arrangements are definitely separated
by great groups, the Cupresseae and Gnetales being c^Jic, the
other Gymnosperms spiral.

The tendency to heterophylly is a very strong one, scale
leaves being associated with foliage leaves in every great group.
In Cycadales there is a regular alternation of the tAvo types,
while in Conif erales the extreme of diversity is reached. In
Coniferales the primitive type of shoot seems to have consisted
of axes all of which were clothed with free needle leaves, which
later became intermixed with scales; while in the highly spe-
cialized genus Pinus the foliage leaves abandon the main axes,
and are pushed out on dwarf shoots to the extremities of the
branch system^

The foliage leaves are remarkably varied in form and size
and venation, being pinnately branched and more or less dichot-
omously veined in Cycadales and Bennettitales, broad and dichot-
omously veined in Ginkgoales, characteristically acicular in most
Coniferales, ribbonlike and parallel-veined in Cordaitales and
Tumboa, and broad and reticulately veined in Gnetum. It is
also a noteworthy fact that the fernlike leaves of some of the
Cycads also show the characteristic circinate vernation of Ferns.

Although variable in form and venation, the foliage leaves
of Gymnosperms are fairly constant in certain points of struc-
ture, all of them, with few exceptions, developing rigid pro-
tective structures, and having a limited distribution of the vas-
cular system, with a characteristic development of " transfusion
tissue."

THE ROOT

So far as studied, the roots of Gymnosperms seem to be of
very uniform structure. As distinct from Pteridophytes the

156 MORPHOLOGY OF SPERMATOPHYTES

strong tap root is to be noted ; while the peculiar origin of the
root cap seems to distinguish the group as a whole from Angio-
sperms. There is no distinct calyptrogen, but the periblem
layers are continuous over the tip, and according to Strasburger
each with its own meristem. As the layers over the tip continue
to divide, the outer ones become more or less separated and
dead, and function as a cap. The primary vascular cylinder
seems to be very uniformly diarch, a simpler type than prevails
among Dicotyledons, and much simpler than in Monocotyledons.

II. THE SPORE-PRODUCING MEMBERS

THE MICKOSPOKANGIUM

The microsporangia are always developed in connection with
a strobilus or strobiluslike structure, which in the genus Cycas
is terminal upon the main axis, and may be so in the other genera
of Cycadales; but it is lateral in all other Gymnosperms, and
either axillary or borne upon axillary spur shoots. The stro-
bili are almost universally monosporangiate, Tumboa giving
evidence of a former bisporangiate condition in the occurrence
of a sterile ovule in the staminate " flowers/7 and the Bennetti-
tales, in some cases at least, displaying a bisporangiate strobilus.

In most Gymnosperms the microsporangia seem to be dis-
tinctly foliar, being borne by an undoubted sporophyll. In
some genera of Cycadales (as Cycas, Macrozamia, Ceratozamia,
and 8tangeria) , and in Bennettitales, so far as known, the very
numerous sporangia, often in definite sori, and borne upon the
under surface of broad scales, are very suggestive of Filicales.
In other Gymnosperms, however, the sporophylls are more modi-
fied, and the sporangia are much fewer in number, frequently
being reduced to two upon each sporophyll. In certain genera
of Cycads (as species of Zamid) and of Conifers (as Taxus\ the
peltate type of sporophyll is found, and the sporophyll of Giiikgo
may be regarded as a modification of it. In Conifers in general,
however, the expanded portion of the sporophyll develops in the
plane of the stalklike base.

In Ephedra and Gnetum the microsporangia seem to be dis-
tinctly cauline, and are developed near the apex of a short lat-
eral axis which bears coalescent bractlets ("perianth"); but
in Tumboa there seems to be no question as to the sporophyll

COMPARATIVE MORPHOLOGY OF GYMNOSPERMS 157

character of the sporangium-bearing structure, which usually
consists of six monadelphous stamens. Of course it is possible to
conceive of the so-called short axis in Ephedra and Gnetum as
monadelphous stamens, but this hypothesis does not seem to be
necessary.

Special interest attaches to the sporangium-bearing structure
of the Cordaitales, both on account of its peculiarity, and on
-account of the antiquity of the group. It is a debatable question
whether the structure is a sporophyll or a lateral axis, but in
any event the sporangia are erect and in a terminal cluster.
The terminal position of the sporangia finds its parallel in all
the genera of Gnetales, and is no indication of their foliar or
cauline character. It seems that the most primitive type of
microsporophyll among Gymnosperms must be that displayed
by Cycas and its related genera, and by the Bennettitales, but
the oldest microsporophylls known, those of Cordaitales, seem to
hold no clear relation to them. It may be that when the spore-
producing members of the Cycadofilices are known some light
will be thrown upon the evolution of the microsporophyll in
Gymnosperms.

The development of the sporangium in all the forms investi-
gated is quite uniform, and is of the ordinary eusporangiate
type. The archesporium consists of a hypodermal plate of
cells, which divide by periclinal Avails, separating an outer and
sterile wall layer from an inner sporogenous one. By subse-
quent divisions a wall of three to seven layers is developed out-
side of the sporogenous tissue and within the epidermis. Nu-
merous spore mother cells are organized, and in this condition
the sporangium passes the winter or the rest period.

Gymnosperms seem to be universally anemophilous, al-
though it is claimed that Tumboa is probably entomophilous.
The extreme of anemophilous adaptation has been attained by
the winged pollen grains of the dominant modern genera Pinus
and Podocarpus.

THE MEGASPOKANGIUM

The presence or not of a megasporophyll is one of the prob-
lems of the group, which has been discussed sufficiently else-
where. It seems clear that the ovules are both foliar and cau-
line. They are certainly foliar in most if not all of the Cycad-

158 MORPHOLOGY OF SPERMATOPHYTES

ales, and probably so in Ginkgo, although in the latter case the
supposed carpel is a very rudimentary structure. Among the
Cordaitales, Bennettitales, Gnetales, and Taxaceae, the ovules
seem to be clearly cauline; while among the Pinaceae the carpel
is said to be " reduced to the ovule " ; the latter therefore would
be essentially cauline. One is tempted to regard the foliar
ovule as the more primitive type, and to trace the reduction
of the sporophyll from Cycas, where it is well expressed, through
Ginkgo, in which it is much reduced, to the Conifers and
Gnetales, in which it is eliminated. To such a view, how-
ever, the testimony of history is opposed, for among the Cor-
daitales, the most ancient of recognized Gymnosperm groups,
the ovules were plainly cauline, and apparently so among the
Bennettitales, the most ancient of the Cycadean forms. The evi-
dence seems to be that the foliar ovule does not necessarily indi-
cate a more primitive group than the cauline, and that the dis-
tinction between these two types of ovules is more arbitrary
than real, often impossible to apply, and of no special signifi-
cance.

The question of integuments is another Gymnosperm prob-
lem. The usual statement is that the integument is single in
all Gymnosperms except Gnetum, and this certainly expresses
the result of ordinary observation. If the definition of integu-
ment be restricted to such structures as receive the name in
Angiosperms, then it can be claimed that one integument is the
rule among Gymnosperms. If it be allowed, however, that the
integument may become so modified as to lose its character of
an ordinary investment of the ovule, we are launched into a dis-
cussion whose merits can not be decided by demonstration. It
is not clear why it is essential for so ancient and diversified
a group as the Gymnosperms not to vary in the number of its
integuments.

As opposed to the ordinary view that one integument char-
acterizes the group, and Lotsy would even include Gnetum in
this generalization, is the view of Celakovsky that the whole
group is characterized by two integuments, the outer one of
which is often remarkably modified. This view certainly
sweeps into one category a number of problematical structures
and has the attractiveness of uniformity. The application of
this theory would result somewhat as follows: Among the an-

I

COMPARATIVE MORPHOLOGY OF GYMNOSPERMS 159

cient Cordaitales the most obvious explanation of the ovule sec-
tion is the presence of two distinct integuments, and the testa
of the Bennettitales indicates the same fact in that group. In
Cycads, Ginkgo, Cephalotaxus, Podocarpus, etc., the outer
fleshy and inner bony layers may be regarded as representing
two integuments structurally, which have become connate;
while in Taxus, Dacrydium, Microcachrys, etc., the outer flesh-
forming integument has remained or has become distinct as the
so-called '" aril." In the Pinaceae the most remarkable modifi-
cation is displayed, the outer integument having entered into the
structure of the " ovuliferous scale." Among the Gnetales the
inner integument is produced into the remarkably long and tubu-
lar micropyle, while the outer one has been usually designated a
" perianth."

It is evident that between distinct bracts on the one hand
and a genuine integument on the other there may be all sorts of.
intergradations, and that a hard-and-fast morphological line
probably does not exist. We can recognize in all Gymnosperms
a single undoubted integument, but whether the next outer
structure should be called an outer integument, an aril, a peri-
anth, an ovuliferous scale, or what not, seems impossible and
perhaps unimportant to decide.

The nucellus of the group presents some features which seem
to be important. In the whole group there is a remarkable de-
velopment of sterile tissue above the single or several sporoge-
nous cells, which thus appear imbedded in the chalazal region of
the ovule. Among the more primitive Gymnosperms (Cor-
daitales, Bennettitales, Cycadales, and Ginkgoales) this sterile
mass differentiates into a firm persistent beak at the apex, and a
loose tissue between the beak and the embryo sac. In the same
groups a deep pollen chamber is developed within the beak, the
loose tissue becomes broken down by rhizoidlike pollen tubes,
and the beak settles down upon the embryo sac. This uniform
structure in all the more primitive forms certainly has to do with
the more primitive phases of siphonogamy, which will be dis-
cussed more fully under fertilization. Among the Coniferales
and Gnetales the beak is not developed, but the outer micropylar
layers of the nucellus are of firmer tissue than those beneath,
and occasionally the tip of the nucellus breaks down into a cup-
like depression which holds the pollen grains, as has been ob-

160 MORPHOLOGY OF SPERMATOPHYTES

served in Pinus and in Gnetum. An exception must be made
of Tumboa, in which the persistent beak is formed, which is
riddled with passageways but contains no definite pollen cham-
ber. Sequoia also presents an' interesting modification, the
micropylar sterile region of the nucellus being but feebly devel-
oped, the sterile tissue beneath the sporqgenous mass growing
in a remarkable way and being penetrated by the elongated and
tubelike niegaspore.

The origin of the deep-seated sporogenous cell or group is
an open question for most forms. The general statement is that
,it is derived from an archesporium consisting of hypodermal
cells, .and that periclinal divisions cut off outer sterile cells
which build up the 'remarkable mass of sterile tissue. This has
been definitely observed in a very few forms and has been in-
ferred for the rest. It coincides with the well-known sequence
in Angiosperms, and may prove to be true of Gymnosperms in
general, but we have not been able to verify it as yet in Pinus.
The first differentiation of sporogenous tissue usually visible is
the appearance of one or several sporogenous cells in the chalazal
region of the ovule.

III. THE GAMETOPHYTES

THE FEMALE G AMETOPH YTE

With the exception of Tumboa and Gnetum, the structure
of the female gametophyte is very uniform in Gymnosperms,
and recalls in a general way the female gametophyte of such
heterosporous Pteridophytes as Selaginella and Isoetes. This
resemblance, however, is very general, and should not be pushed
too far, as is shown by recent investigations.

The general statement is that the mother cell divides and
organizes a row of potential megaspores, the lowest of which
becomes the functional niegaspore. We have recently succeeded
in verifying this statement for Pinus (Fig. 106). In this genus
we have repeatedly observed a single cell, usually centrally placed
in the chalazal region, which by its increase in size and its con-
tents is evidently the functional mother cell ; but we have only
a single preparation to show that it develops a row of potential
megaspores. We suspect that in Pinus, and in many other
Gymnosperms, there is but a single primary sporogenous cell,

COMPARATIVE MORPHOLOGY OF GYMNOSPERMS

161

which passes over directly into a mother cell, and that the pecul-
iar differentiation and breaking down of a zone of tissue about
it has led to the impression of a sporogenous mass.

In some cases where there is more than one mother cell, two
or more megaspores begin to develop, and in Sequoia the sterile
mega spores are prominent and persistent, but as a rule one soon
becomes dominant at the expense of the rest and of the adjacent
nucellar tissue.

The mother cell increases in size in the usual way, and early
in its history begins to germinate. The sequence of events
seems to be very uniform, being free nuclear
division (usually about eight simultaneous
divisions), parietal placing of free nuclei in
a cytoplasmic layer, organization of pari-
etal tissue, and gradual filling of the embryo
sac by centripetal growth. The cases of
Tumboa and Gneium are peculiar in that
there is a distinct polarity in the tissues of
the sac, and an appearance of reproductive
cells in a sac not filled with tissue. In Turn-
boa the micropylar region of the sac con-
tains cells but loosely aggregated when the
reproductive cells appear, while in Gnetum
the micropylar end of the sac, and in some
species the whole sac, contains only free
nuclei when fertilization occurs. Indica-
tions of the same tendency in other forms
are not lacking. In Sequoia it is reported
that while cell formation is taking place
at the two extremities of the sac, free nuclear division con-
tinues in the middle region, the sac thus being distinctly divided
into its reproductive and nutritive regions.

In the development of the archegonia, which are variable in
number and in distribution, there are some features that seem
peculiar to the group. The neck canal cell 'series, which is by
no means prominent in the Pteridophytes, is not represented at
all in the Gymnosperms; and the ventral canal cell, cut off
immediately before fertilization, is usually very minute and
ephemeral. Probably the most noteworthy fact in connec-
tion with the archegonium is the organization of a sheath of

FIG. 106. — Pinus Lari-
cio, showing the four
potential megaspores
derived from the
mother cell, x 666.

162 MOKPHOLOGY OF SPERMATOPHYTES

jacket cells about the central cell, which through numerous wall
pores empty their contents, nuclei and all, into the rapidly
enlarging central cell, and this material remains conspicuous
in the large egg. The neck is quite variable in structure,
organizing from two cells to several tiers with eight cells in
each tier. Probably the most common structure is two tiers
with four cells in each tier. The two-celled neck, so far as re-
corded, occurs in Cycadales, Gmkgoales, Tsuga Canadensis, and
Cephalotaxus Fortunei. Tumboa and Gnetum develop no arche-
gonia, as would be expecied from the nature of their endo-
sperm tissue. In Tumboa the archegonium initials are selected,
but no division occurs, the egg organizing directly in the initial ;
while in Gnetum the eggs are as free and naked as in the Angio-
sperms. The last two forms are particularly suggestive in
representing the possible method of transition from a compact
ante-fertilization endosperm to one that consists of free nuclei
when fertilization occurs, a very characteristic feature of Angio-
sperms.

THE MALE GAMETOPHYTE

While information concerning the germination of the micro-
spores of Gymnosperms is still very much needed, enough has
been observed to indicate the general character of the game-
tophyte. In most cases one or two vegetative (" prothal-
lial ") cells are first cut off. In Cycadales a single such cell
is cut off and persists ; in Ginkgoales and Ephedra two such
cells are successively cut off, the first being ephemeral ; while in
Abieteae two cells are cut off, and both are ephemeral. In cer-
tain forms studied, as Taxus, Juniperus, Cupressus, and Se-
quoia, no evidence of vegetative cells has been obtained. It
should be remembered, however, that these cells are cut off rap-
idly and are very ephemeral, so that traces of them may be lost
quickly, or they may easily escape observation. We venture
the opinion, therefore, that one or two vegetative cells, more or
less evanescent, will be found to be of common occurrence among
Gymnosperms — cells which are alm6st unknown among Angio-
sperms.

The large cell which remains after the cutting off of the
vegetative cells we have taken to be the morphological equivalent
of the antheridium initial as it appears in the heterosporous

COMPARATIVE MORPHOLOGY OF GYMNOSPERMS 163

Pteridophytes, although we recognize the fact that there may
be grounds for considering the generative cell as the antheridium
initial, in which case the structures can not be homologized with
those of the heterosporous Pteridophytes. This large cell di-
vides unequally, giving rise to a sterile tube cell (the larger),
which is concerned with the development of the pollen tube, and
a smaller primary spermatogenous cell or generative cell. The
spermatogenous series, beginning with the generative cell, seems
to be very uniform in Gymnosperm^, and is longer by one gen-
eration than is the series in Angiosperms. The generative cell
divides and forms the stalk and body cells, the former of which
functions no further. The body cell divides and organizes the
two functional male cells, the morphological equivalents of the
sperm mother cells of Pteridophytes. If the stalk cell, sister of
the body cell, should also divide,, four male cells would result,
and the sequence would be that of Isoetes. In the Cycads these
male cells become ciliated, through the very peculiar activities
of blepharoplasts, but they should not be regarded as the homo-
logues of the multiciliate sperms of Pteridophytes. Whether
the ciliated sperm of Giiikgo is the transformed mother cell or
not is a matter of doubt, for although the details of its forma-
tion seem to be identical with those of the ciliated male cells
of the Cycads, there is evidence that they are organized within
the mother cells and discharged, in which case they are the mor-
phological equivalents of the sperms of Pteridophytes. It is
barely possible that there is here a transition form between a
discharged ciliated sperm and a retained sperm resulting in a
ciliated mother cell.

Some meager evidence as to the nature of the male gameto-
phyte of Cordaitales is very suggestive. Within the well-pre-
served pollen grains a mass of tissue has been observed, whose
nature, whether vegetative or antheridial, it is impossible to
determine. In either event it indicates a much more primitive
male gametophyte than has been found in any living Gymno-
sperm. The testimony from the structure of the male gameto-
phyte indicates an origin of Gymnosperms from forms whose
gametophytes had more complex vegetative tissue than is shown
by any living heterosporous Pteridophyte.

164: MORPHOLOGY OF SPERMATOPHYTES

FERTILIZATION

It is now known in Cycadales and Ginkgoales, forms with
ciliated male cells or sperms, that these cells are not trans-
ferred through the pollen tubes. Branching tubes penetrate in
every direction through the loose nucellar tissue which lies be-
tween the beak and the embryo sac, and seem to be purely absorp-
tive in function. As the nucellar tissue is broken down, the
pollen grains containing the body cell are brought into close
proximity to the archegonia, so that when the ciliated male
cells or sperms are organized and discharged they are able to
reach the necks by swimming. As the Cordaitales and Ben-
nettitales show nucellar structures identical with those of Cyca-
dales and Ginkgoales, it seems reasonable to suppose that their
pollen tubes, if any existed, were also rhizoidal in character,
and that their sperms were ciliated. The conclusion seems to be
inevitable that pollen tubes appeared primarily as absorbing
organs, and that their function as sperm carriers is one that
should be regarded as secondary.

In the other groups of Gymnosperms, so far as known, the
pollen tube, while retaining more or less of its rhizoidal char-
acter, is also distinctly a sperm carrier, and this function seems
to be associated with the development of nonciliated male cells.
In these cases the whole of the contents of the swollen tip of the
tube, consisting prominently of the two male cells, tube nucleus,
and stalk-cell nucleus, has been observed to be discharged into
the cytoplasm of the egg.

The development of two male cells seems to be thoroughly
established throughout the Spermatophytes. In the grouped
archegonia of Cupresseae both male cells may function, and the
same may be true in the case of branching tubes which are sperm
carriers. In Taxus, however, the inequality of the male cells
indicates that one of them has long since ceased to be functional.

The discharge of the contents of the pollen grain or pollen
tube is in every case closely preceded by the cutting off of the
ventral canal cell. Occasionally the nucleus of this cell becomes
as highly organized as that of the egg, and there is reason to
believe that in some instances it may be fertilized. The cyto-
plasm of the egg receives the various contents of the grain or
tube, but everything remains near the periphery of the cyto-

COMPARATIVE MORPHOLOGY OF GYMNOSPERMS 165

plasm except the nucleus of the male cell, which advances toward
that of the egg and becomes imbedded in it. There is some
evidence that the fusion of the two nuclei does not involve a
fusion of their chromatin, and that the male and female chroma-
tin continues to exist as separate groups at the first segmenta-
tion of the fusion nucleus. How far this may be true of Gym-
nosperms in general it is impossible to say, but in any event it
is an interesting problem, involving not only a departure from
the habit of other known plants, but also a resemblance to the
phenomena of fertilization observed in certain animals, as
Ascaris.

It is perhaps well to call attention to the fact that the cyto-
plasm of the male cell enters the cytoplasm of the egg and fuses
with it. The importance of this in connection with fertiliza-
tion is a matter of opinion rather than of demonstration. It is
certainly true that the nucleus of the male cell of Gymnosperms,
and probably of all plants, becomes relatively very large, being
covered by a thin layer of cytoplasm. Whether this cytoplasm is
directly necessary to fertilization, or merely necessary to the male
nucleus preceding fertilization, is a question for the future to
answer.

IV. THE EMBRYO

The germination of the oospore is one of the most character-
istic features of Gymnosperms, and although the details seem to
vary widely there is underneath them all a general method which
belongs to the group and has been observed in no other. A
natural series may be arranged, beginning with the Cycadales.
In them, germination begins with free nuclear division, fol-
lowed by the placing of the free nuclei in a cytoplasmic layer
which lines the wall of the oospore and surrounds a central cav-
ity. The free nuclei then organize a parietal layer of cells,
which are apt to be more or less massed at the base of the oospore
(towards the main mass of the endosperm), the whole of the endo-
sporic tissue being known as the proembryo. These basal cells
of the proembryo then begin a remarkable elongation, forming
the very long and massive suspensor, which bears at its tip the
group of cells which is to develop into the embryo proper. The
three regions of the embryo which thus appear are (1) the pro-
embryo, an endosporic structure, which relates the embryo to

166 MORPHOLOGY OF SPERMATOPHYTES

the nutritive supplies stored up in the egg, (2) a suspensor, which
develops from the proembryo and relates the embryo to the
nutritive supplies of the endosperm, and (3) the embryo proper.

In the case of the Coniferales the same sequence is repeated,
except that free nuclear division becomes much reduced, usually
resulting in but four nuclei, which pass to the base of the oospore
and organize a definite basal proembryo of three tiers of cells,
with four cells in each tier. It is interesting to note that in
Cephalotaxus Fortunei the free nuclear division is not so re-
stricted, but results in eight or sixteen free nuclei, which pass
to the base and organize the proembryo. The difference be-
tween Coniferales and Cycadales, therefore, is that in the for-
mer the number of free nuclei is restricted, and therefore the
formation of a proembryonic tissue takes place only in the basal
region of the oospore. In both cases the suspensor is formed by
the elongation of some of the cells of the proembryo, and in
Coniferales a single plate ("rosette") of cells retains connec-
tion with the nutritive supplies of the oospore, rather than a
parietal layer. Another feature in the formation of the pro-
embryo of Coniferales is worthy of note. The first walls formed
about the nuclei of the primary basal group are not an accom-
paniment of nuclear division, but seem to be formed directly by
fibers investing the nuclei, and they do not cut off these nuclei
from the general cavity of the oospore above. The subsequent
walls which result in building up the tiers are associated with
nuclear division, but the uppermost nuclei always remain ex-
posed to the cavity of the oospore, and are separated from one
another by walls which extend more or less into the mass of the
oospore.

In Ephedra free nuclear division also occurs, two to eight
such nuclei being formed, but they organize as free cells, and
do not form a definite tissue. As a consequence, the cells act
independently, and each one develops a long tubular process
(suspensor), the bulbous base remaining in touch with the nutri-
tive supplies of the spore, and at the tip of each suspensor an
embryonic cell is cut off. This independence of the cells may
be observed also in Pinus, where, although there is a regular
tier formation of the proembryo, each cell of the suspensor
set develops a separate suspensor bearing at its tip an em-
bryonic cell.

COMPARATIVE MORPHOLOGY OF GYMNOSPERMS

167

In Tumboa and Gnetum, free nuclear division, which in
Ephedra sometimes results in hut two nuclei, is abandoned en-
tirely, and the whole oospore behaves as do one of the free cells
in Ephedra, elongating in suspensor fashion, and cutting oif at
the tip an embryonic cell.

The case of Ginkgo is a very interesting departure from the
ordinary sequence. The germination of the oospore begins with
free nuclear division, but suspensor development is entirely
omitted, and as a consequence the oospore is filled with a com-
pact tissue which organizes directly into the embryo, and this
encroaches upon the endosperm by the growth of the whole
mass. It is important to note that the method of Ginkgo was
also that of the Bennettitales, and this fact raises the question
whether this does not represent the type of embryogeny among
the primitive Gymnosperms, in which the three embryonic re-
gions of modern Gymnosperms have not been differentiated.
Xo evidence as to the embryo of the Cordaitales has been ob-
tained, so that the very important testimony which they might
give is not available. If this be the more primitive type of
embryogeny, it would follow that the conspicuous suspensor of
most living Gymnosperms is a structure of secondary origin in
the group. However, these two types of embryogeny may not
hold any definite relation to each other.

The development of the embryo proper seems to follow no
rigid sequence. The first division may be transverse or vertical,
and the subsequent divisions may be in almost any order or
direction. Perhaps too much emphasis has been laid upon the
sequence of events in the earliest stage of the embryo ; at least,
the data now at hand seem to indicate the entire lack of a definite
sequence. A study of the organization of the growing points of
the great body regions, however, should lead to important re-
sults, but unfortunately the knowledge of Gymnosperms in this
regard is too meager to justify any definite statement.

It is evident that polyembryony is natural in the group, and
numerous embryos in various stages of development are com-
monly found imbedded in the endosperm, but for more than one
embryo to come to maturity seems to be a rare phenomenon.

12

CHAPTER VII
THE PHYLOGENY OF GYMNOSPERMS

ANY statement as to the phylogeny of a group must be largely
hypothetical. The data upon which opinions are based are
never sufficient, but such opinions serve to coordinate knowl-
edge and to suggest profitable research. Only a small fraction
of living forms have been studied adequately ; while the extinct
forms, which represent the early history essential to phylogeny,
will probably never be known except in an uncertain and frag-
mentary way. In the case of Gymnosperms we have the advan-
tage of dealing with a woody group, whose fossil remains may be
taken to represent their ancient history with unusual complete-
ness. But even in the presence of an abundance of remains, the
facts of greatest service to morphologists are for the most part
inaccessible.

Before suggesting a possible phylogeny for the group and
its various members, it may be well to state certain factors
which must enter into any consideration of such a subject. In
the first place, it should be remembered that it is exceedingly
inipr^bable_that_any important group of living^ forms .hasjbe_en
derived from anotErgroiTpIHIiESTig -Fnrrng" Resemblances in
structure which are regarded as essential may be pointed out,
but this is not likely to mean the origin of one group from the
other; it may mean that the two groups can be traced to one,
probably now extinct, which combines the characters now dif-
ferentiated. Most living groups are best regarded as divergent
rather than consecutive series.

Another important pointjsJJial similaj^chariges in structure
may have appeared independently in different lines! The re-
sponse oTofgamsms in their structure to environment is deeper
seated than we were once inclined to believe, and testimony
from the similarity of certain structures, when contrary to the
168

THE PHYLOGEXY OP GYMXOSPERMS 169

testimony of the majority of other structures, argues feebly for
recent community of origin. For example, we are compelled to
believe that so important a condition as heterospory was attained
independently by several lines. To put into the same genetic
group all heterosporous Pteridophytes would be regarded as a
morphological absurdity. If heterospory appeared independ-
ently in several lines, the same conclusion must be reached in
reference to its natural outcome, the seed, and the polyphyletic
origin of the Spermatophytes becomes extremely probable. A
corollary of the point last mentioned is that a living group may
not have a common phylogeny with any other living group show-
ing similar structures. For example, we are not necessarily re-
stricted to the heterosporous Pteridophytes in searching for the
ancestral forms of the Gymnosperms, for the latter may repre-
sent a distinct heterosporous line. While this increases the per-
plexities of phylogeny, it broadens its horizon, and homosporous
Pteridophytes may be included in the search for the origin of
the Gymnosperms.

What seems to be a conspicuous error in many, schemes of
phylogeny is the tendency to focus attention upon very few
structures. It may be that the structures selected are the most
significant in a general way, but the organism is a plexus of
structures and must be considered in its totality. Very differ-
ent structures must have been laid hold of by the processes of
evolution, and it may not be possible to relate the resulting forms
properly upon the basis of any one or two structures. With all
of these uncertainties in mind, we may reach not a clear phylog-
eny of the Gymnosperms, but a clearer understanding of the com-
plexity of the problem and of the uncertainty of conclusions.

Ever since Hofmeister's classic researches, it has seemed
clear that the Gynmosperms have been derived from Pterido-
phyte stock. As this view meets general consent there is no
need to discuss it. It was also natural for a time to regard
Gymnosperms as phylogenetically intermediate between Pter-
idophytes and Angiosperms, for it was not easy to believe that
such a structure as the seed could have appeared in more than
one genetic line. It is probably not going too far to say that
there is now no serious opposition to the view that the (rypiTiff-
sperm and Angiosperm lines are genetically independent. The
problem, therefore, has to do with the relation of the Gyinno-

170 MORPHOLOGY OF SPERMATOPHYTES

sperms and the Pteridophytes to one another, and does not con-
cern itself in any way with the Angiosperms.

Assuming the origin of Gymnosperms from Pteridophytes
as true, the question is as to the monophyletic or polyphyletic
origin of the group. Was a single group of archaic Gymno-
sperms derived from Pteridophytes, which subsequently differ-
entiated into distinct lines; or have several Gymnosperm lines
originated independently from the Pteridophyte stock? The
great diversities which exist among the living representatives of
the group suggest a p^l^phyle^tic_^rigin^ but the numerous im-
portant structures in common, and more than all the testimony
of the great Gymnosperm plexus of the Paleozoic, seem to argue
far more strongly for a monophyletic origin. The reasons for
these views will become more clear as the interrelationships of
the groups are considered.

Assuming the monophyletic origin of Gymnosperms, the
question is at once suggested as to the special group of Pterido-
phytes which gave rise to them. Curiously enough, each of the
three living phyla of Pteridophytes has been claimed as having
been the source of the Gymnosperms. The hypothesis which
looks to the Calamodendreae among the Equisetales as the an-
cestral forms has now few if any supporters. A persistent
hypothesis, however, associates Gymnosperms with the Lycopo-
diales, through such ancient forms as the Sigillarieae, an origin
whicH is claimed for the Coniferales, even when the Cycads are
acknowledged to be of filicinean origin. A third hypothesis
would derive the Gymnosperms from the Filicales. As Pro-
fessor Scott has said : * " These interpretations, whether right
or wrong in the particular cases, at least indicate the broad fact
that important anatomical characters, which \ve are accustomed
to associate with Gymnosperms and dicotyledonous Phanero-
gams, were in Paleozoic times common to a large proportion of
the vascular cryptogams. Community of secondary tissue for-
mation, however, is by itself no proof of affinity, and the problem
must be attacked on other lines."

The view which we maintain is that the Cycadales have un-
doubtedly been derived from the Filicales, and that it is impos-
sible to dissociate the other Gymnosperm lines from the Cycads.

* Studies in Fossil Botany, p. 513, 1900.

THE PHYLOGENY OF GYMNOSPERMS

171

In the present state of knowledge, it is scarcely necessary to
recapitulate the arguments which favor a filicinean origin for
the Cycads, especially since this view seems to meet with general
consent.

Approaching the subject from the historical standpoint, we
find that the Filicales represent an extremely ancient line, prob-
ably the most ancient of the Pteridophyte lines, reaching back
in history at least into the Silurian^ The group is such an
ancient one that in all of its known history it seems to be en-
tirely independent of the other series of Pteridophytes. In the
Carboniferous the display of Filicales was an extraordinary one.
Associated with true Ferns was an almost equally extensive
group which the paleobotanists have called Cycadofilices. The
latter forms so much resembled the Ferns in outward aspect
that they Avere originally all included among the Ferns; but
subsequent study has revealed the fact that their anatomical
structure shows a remarkable commingling of Fern and Cjcad
characters. Unfortunately, the sporangia_^of Cycadofilices are
unknown, and the group must at present remain .of indefinite
limitation. The main fact, in the present discussion, however,
is that a very extensive group, showing characters intermediate
between Ferns and Cycads, existed during the Carboniferous
period in connection with undoubted Ferns. From the great
antiquity of the Filicales, the conclusion seems to be rational
that during the great vegetative display of the Carboniferous
there was a differentiation of the phylum, which gave rise to the
Cycadofilices, the old Fern stock continuing to display itself in
forms which would be recognized to-day as members of the
Marattiaceae. The first conclusion, therefore, would be that
the Gymnosperm phylum began with the separation of the
Cycadofilices from ancient marattiaceous Filicales.

Associated with the Cycaclp^li££S_diiring the Carboniferous
was the greaFgroup Cordaitales, undoubted Gymnosperms, and
related to both Cycads and Ferns in such a way as to suggest
very definitely an origin similar to that of the Cycads. Whether
the Cordaitales came directly and indejjgjidejiiLy from the Fili-
cales or not is doubtful, but our judgment favors the latter view,
since we are inclined to regard the Cycadofilices as represent-
ing the original indefinite plexus out of which all the Gymno-
sperm lines have been derived.

172

MORPHOLOGY OF SPERMATOPHYTES

During the Carboniferous, also, forms appeared which seem to
be referable distinctly to Ginkgoales, a line which in its begin-
nings has seemed hard to distinguish from certain of the Cor-
daitales. Ginkgo has always been peculiar in its combination
of Cycad and Conifer characters, and this may be accounted for
by its early origin in connection with the Cycadofilices and Cor-
claitales.

During the Mesozoic there appeared the Bennettitales and
Cycadales, the former making much the greater display. It
is reasonable to suppose that these two lines were derived either
independently or together from the Paleozoic Cycadofilices.

The historic evidence concerning the Coniferales is most
indefinite and uncertain. The group, as we understand it to-
day, is relatively modern, but there is no evidence as to the rela-
tive ages of the two great lines, Taxaceae and Pinaceae, which
compose it. For a long time the genus Araucaria has been
regarded as the most ancient representative of the Conifers, but
this idea seems to have arisen from the Araucaria-like wood
which is so abundant in the Paleozoic, which has proved to be-
long to the Cordaitales. Just which of the Conifer lines is
the most ancient remains for future investigation to discover;
that the line as a whole was an offshoot from the Cordaitales
seems to us most probable. In fact, the Cordaitales, as ordi-
narily recognized in the Carboniferous,
seem to us to be forms well on their way
toward the Coniferales.

In reference to Gnetales, there
seems to be absolutely no historic evi-
dence, and morphology is equally at
fault. That the three genera now ex-
isting are a fragmentary representation
of some ancient line seems to be a rea-
sonable conclusion; and the characters
are related so distinctly to those of other
Gymnosperms that it seems hardly a

question but that the group has the same origin, although one can
not venture upon any more detailed hypothesis.

In order to summarize clearly the statements made above,
the accompanying diagram is introduced, but it must be inter-
preted in a most general sense.

THE PHYLOGENY OF GYMNOSPERMS 173

The origin of both Gymnosperms and Angiosperms is bound
up with the evolution of heterospory and the seed habit, a state-
ment concerning which is presented in the following quota-
tion : * " The evolution of heterospory seems simple enough.
The physiological differentiation of the spores was complete
when prothallia became persistently dioecious. This division
of labor is to be expected in the case of two such distinct func-
tions as the production of antheridia and of archegonia. A
prothalliuni producing both sex organs equally well may be
regarded as in a state of equilibrium, an equilibrium which is
disturbed by any conditions which favor the production of one
sex organ rather than the other, in this case probably nutritive
•conditions. This disturbance of the equilibrium of a bisexual
prothalliuni would certainly find expression first in a dioecious
tendency, and finally in a dioecious habit. With the habit
once fixed the morphological differentiation of spores becomes
inevitable, since the nutritive requirements of the two pro-
thallia are so different. The evolution of heterospory seems
to be one of the simplest of selective processes, with inequalities
of nutrition to furnish the variations. . From this point of view
it would seem natural to expect that it may have been derived
frequently from homospory.

" The retention of the megaspore, however, does not seem to
be so simple a problem. In a certain sense it is correlated with
the reduction of the gametophyte, since retention would not
seem practicable until reduction had proceeded far enough to
make the gametophyte endosporic. Even greater reduction,
however, is attained by the male gametophyte, but the spore is
shed. It should be noted that even in the case of the microspore
the male gametophyte is usually completely organized before
pollination; but the fact remains that the reduction does not
compel retention.

" It has seemed to me that this phenomenon is to be ex-
plained by the law of progressive sterilization. This law cer-
tainly finds expression in the megasporangia of heterosporous
Pteridophytes, in which the sterilization of mother cells is con-
spicuous. This method of increasing the nutrition of the fertile
cells is too common a phenomenon to need illustration ; but it is

* COULTER, JOHN M. The Origin of Gyranosperms and the Seed Habit.
Bot. Gaz. 26: 153-168. 1898.

174 MORPHOLOGY OF SPERMATOPHYTES

a tendency that would seem very consistent with the develop-
ment of megaspores, whose peculiar work holds so definite a re-
lation to abundant nutrition. For this very reason high num-
bers of microspores may be continued, and a diminishing number
of megaspores produced. This would reach its culmination in
the production of but a single megaspore by a sporangium, and
a proportionate increase in the size of the megaspore. With the
development of a single spore imbedded in a sterile tissue, shed-
ding becomes not only mechanically difficult, but meaningless,
since the necessity for scattering abroad of gametophytes, to
avoid competition, has disappeared. It is further true that the
development of such a spore involves nutritive supplies from
numerous neighboring cells, and a certain amount of retention
becomes necessary for this reason. Still further, the advantage
to a single megaspore in being retained, thus securing more abun-
dant outside nutrition during germination, would fix the habit
if any selective process were at work. For these various rea-
sons it would seem evident that when the sterilization of a mega-
sporangium had reached its extreme limit, by organizing a single
spore, retention is likely to follow sooner or later. If this line
of reasoning be true, the seed habit might have been developed
in any heterosporous line."

Some unpublished results obtained by Miss F. M. Lyon in
a study of species of Selaginella are of interest in this connec-
tion. The usual condition was found in which the one or two
fully formed megaspores were retained in the sporangium until
they germinated. Not only this, but the eggs were fertilized
and complete embryos formed within the retained megaspores.
Fertilization was effected not by the entrance of sperms into the
sporangium, but by the entrance of microspores, which dis-
charged sperms through papillate protuberances. Moreover,
the embryos thus formed escaped from the megaspores and
developed into young plantlets, a strobilus often being cov-
ered with sprouting plantlets. It should be remarked that such
strobili had fallen from the parent plant, so that the megaspores
had been physiologically separated from the individual which
produced them, but the permanent retention of the megaspore
within the sporangium, without interfering with germination,
fertilization, or the development and escape of the embryo, are
facts which contain a morphological suggestion.

CHAPTEK VIII

GEOGRAPHIC DISTRIBUTION OF GYMNOSPERMS

ALTHOUGH geographic distribution is not a part of mor-
phology, the special student of Gymnosperms should be familiar
with the broad outlines of the distribution of the living forms.
Xo attempt will be made to trace the historical migrations of
the different types, or to indicate natural physiographic areas.
It is sufficient to say in general that there is evidence that most
of the forms once had a very much more extended range than
now, and that the change has consisted chiefly in. a gradual re-
striction of the areas occupied.

CYCADALES

At present the Cycads are strictly tropical forms, the nine
genera, containing something less than one hundred species,
being distributed about equally between the oriental and occi-
dental tropics. Cycas, the oldest genus, containing about
sixteen species, is the most widely distributed oriental genus,
ranging throughout tropical Asia, the East Indies, and the Aus-
tralian region; while Macrozamia, with fourteen species, and
the monotypic Bowenia, are strictly Australasian; Encephalartos,
with twelve species, and the monotypic Stangeria, are African.

In the occidental tropics the largest genus is Zamia, with
about thirty species, ranging through tropical and subtropical
America; Ceratozamia, with six species, and Dioon, with two
species, are Mexican ; while the monotypic Microcycas is Cuban.

It may be remarked in general that Cycas and Zamia repre-
sent the typical Cycads of the two hemispheres, while the other
genera represent relatively isolated forms which bear the stamp
of local conditions.

175

176 MORPHOLOGY OF SPERMATOPHYTES

GINKGOALES

This phylum, once with a very extensive range, is now repre-
sented by a single species, which is restricted to China and
Japan. It was formerly supposed that even this species had
long ceased to exist in the wild state, but specimens which seem
to be undoubtedly wild have been reported from China and
Japan.

CONIFERALES

While Cycads are tropical forms, Conifers belong as dis-
tinctly to temperate regions, forming vast forest areas, especially
in the northern hemisphere. A map indicating their general
distribution would show a heavy north temperate massing, and
a lighter south temperate massing, the two separated from one
another by a broad tropical belt, traversed in but two places —
namely, the East Indian region and the Andean region. These
northern and southern masses contain mostly what are regarded
as different generic types, Pinus dominating at the north, and
Podocarpus at the south. Through the East Indian region
Podocarpus reaches China and Japan ; while the northern genus
Libocedrus reaches Australia by the same route, and penetrates
far into temperate South America by way of the Andes. Aside
from these twro lines of communication, and these two genera,
there -is no crossing of the tropics, all other genera being exclu-
sively northern or southern.

Before considering the main groups in some detail, it
may be well to note a few facts of general interest. By far
the greatest Conifer display in genera and species is that which
borders the Pacific Ocean, the chief areas being the China-
Japan region, the Australasian region, and western North
America; while the whole of this vast border region is occu-
pied by Libocedrus.

The most remarkable displays of endemic genera are found
in the China-Japan and the Australasian regions, the former
containing Cephalotaxus (4 spp.), Pseudolarix (1 sp.), Kete-
leeria (2 spp.), Sciadopitys (1 sp.), Cunningliamia (1 sp.),
Cryptomeria (1 sp.), Glyptostrobus (2 spp.), TJiujopsis (1
sp.) ; while the latter contains Microcachrys (1 sp.), Phyllo-
cladus (3 spp.), Agathis (4 spp.), Arthrotaxis (3 spp.), and

GEOGRAPHIC DISTRIBUTION OF GYMNOSPERMS 177

Actinostrobus (2 spp.). It will be noted that these restricted
genera are well distributed among the great groups, and that
many of them are monotypic. It should be remarked that
in the Australasian region are included the principal East
Indian and Polynesian Islands. Many of these genera are
known to have had formerly a very extensive range, but why
the two regions cited, together constituting practically the west-
ern coast region of the Pacific, should have proved so favorable
for the preservation of types otherwise extinct is an interesting
question.

The other regions of endemic genera are ^orth America,
with Sequoia and Taxodium] and South America, with the
monotypic Saxogothaea of the mountains of Patagonia.

Throughout the northern hemisphere the dominant and wide-
ly distributed genera are Pinus (70 spp.), Juniperus (30 spp.),
Abies (20 spp.), Picea (12 spp.), Cupressus (12 spp.), Larix
(8 spp.), and Taxus (8 spp.). There is. also a remarkable pair-
ing of western iSTorth America and eastern Asia in the display
of certain genera, those common to both regions, besides the gen-
era which extend broadly over both hemispheres, being Torreya,
Tsuga, Pseudotsuga, Thuja, Libocedrus, and Chamaecyparis.

The distribution of forms in the southern hemisphere is
modified by the fact that temperate conditions occur in three
great isolated areas, although this has not resulted in as com-
plete a separation of genera as might have been expected.. The
dominant Podocarpus (40 spp.) is the only genus which occurs
in all three of the south temperate regions ; but in the display
of certain other genera there is a pairing of the continents, the
Australasian region always being one member of the pair, and,
with one exception, South America being the other member.
For example, Fitzroya (2 spp.), Libocedrus (8 spp.), Dacry-
dium (12 spp.), and Araucaria (10 spp.) are common to the
Australasian region and South America; while Callitris (15
spp.) is common to the Australasian region and Africa. The
occurrence of Callitris in the Mediterranean region of Africa,
however, suggests a comparatively recent connection with the
Australasian region through the northern border region of the
Indian Ocean.

In considering the distribution of the great groups of Coni-
fers one is struck by their somewhat rigid geographic limitation,

178 MORPHOLOGY OF SPERMATOPHYTES

and by the occasional remarkable exceptions. Six apparently
natural groups are recognized, and each one belongs distinctly
to the northern or southern hemispheres. Taxeae, Abieteae,
Taxodieae, and Cupresseae are distinctly northern types; while
Podocarpeae and Araucarieae are just as distinctly southern.
It is interesting to note that this division disregards the great
division of Conifers into Taxaceae and Pinaceae.

While this is the general fact, it is more or less true in the
different groups. For example, Abieteae are absolutely north-
ern and Araucarieae are absolutely southern, but the same
sweeping statement can not be made in reference to the other
groups. Podocarpeae, characteristically southern, are repre-
sented in eastern Asia by species of Podocarpus. The three re-
maining northern groups, however, show more striking depart-
ures from the rule. Taxeae are represented in the Australasian
region by the endemic genus Pliyllocladus', Taxodieae are rep-
resented in the same region by the endemic genus Arthrotaxis ;
while Cupresseae are represented in the southern hemisphere by
the endemic Australasian genus Actinostrobus, the Australasia-
African genus Callitris, and the Australasia-South American
genus Fitzroya. The occurrence of these widely separated gen-
era, on the hypothesis that they belong to the same natural
group, is an interesting problem.

From the standpoint of isolated genera, however, the Tax-
odieae are most interesting, for all of the seven genera are very
much restricted. Glyptostrobus is Chinese, Sciadopitys is
Japanese, Cunninghamia and Cryptomeria are China-Japanese,
Arthrotaxis is Australasian, Sequoia is Californian, and Tax-
odium is east North American.

The facts presented above can be understood only when con-
nected with the ancient distribution of the various groups, but
unfortunately the history of Conifers is not in a condition to
supply definite testimony.

GNETALES

Ephedra (30 spp.) occurs in the arid regions of Mediter-'
ranean Europe and adjacent Asia, as well as of America, in
tropical, subtropical, and temperate conditions. Gnetum (15
spp.) ranges through the tropics of both hemispheres. The

GEOGRAPHIC DISTRIBUTION OF GYMNOSPERMS 179

monotypic Tumboa is narrowly restricted to certain extremely
arid regions of western Africa.

There is nothing in the geographical distribution of this
group to suggest anything in common, for the range and condi-
tions are as diverse as are the structures. If they are genet-
ically related, they must represent a very ancient and once
widely distributed phylum, but no sure evidence of its presence
has been discovered among fossil plants.

SUMMARY OF LITERATURE CITED

ARNOLD, W. Bull. Soc. Imp. Nat. Moscow. 329-341 pis. 2. 1899
(1900). See review in Jour. Roy. Micr. Soc. 1900 : 428.

ARNOLDI, W. Beitrage zur Morphologic der Gymnospermen. III.
Embryogenie von Cephalotaxus Fortunei. Flora 87: 46-63.
pis. 1-3. 1900.

Beitrage zur Morphologic der Gymnospermen. IV. Was sind die
" Keimblaschen " oder u Hofmeister's-Korperchen " in der Eizelle
der Abietirieen ? Flora 87: 194-204. pis. 6-7. 1900.

BAILLON, H. Recherches organogeniques sur la Fleur femelle des
Coniferes. Ann. Sci. Nat. IV. 14: 186-199. pis. 12-13. 1860.

BECCARI, O. Delia organogenia dei fiori feminei del Gnetum Gnemon
L. Nuovo Giorn. Bot. Ital. 7 : 91-99. 1877.

BELAJEFF, W. Zur Lehre von dem Pollenschlauche der Gymnosper-
men. Ber. d. deutsch. bot. Gesell. 9 : 280-286. pi. 18. 1891.
Zur Lehre von dem Pollenschlauche der Gymnospermen. Ber. d.
deutsch. bot. Gesell. 11 : 196-201. pi. 12. 1893.

BLACKMAN, V. H. On the Cytological Features of Fertilization and
Related Phenomena in Pinus silvestris L. Phil. Trans. Roy.
Soc. 190: 395-426. pis. 12-14. 1898.

BOWER, F. O. On the Germination and Histology of the Seedlings of
Welwitschia mirabilis. Quart. Jour. Micr. Sci. 21 : 15-30. pis.
3-4. 1881.
On the Further Development of Welwitschia mirabilis. Quart.

Jour. Micr. Sci. 21 : 571-594. pis. 22-23. 1881.
The Germination and Embryology of Gnetum Gnemon. Quart.
Jour. Micr. Sci. 22: 278-298. pi. 25. 1882.

BRAUN, A. 1853.

BRONGNIART, A. Recherches sur 1'Organisation des Tiges des Cyca-t
dees. Ann. Sci. Nat. I. 16: 389-402. pis. 20-22. 1829.

BROWN, ROBERT. Character and Description of Kingia, a new Genus
of Plants found on the S. W. Coast of New Holland, with Obser-
vations on the Structure of its Un impregnated Ovulum and the
Female Flower in Cycadeae and Coniferae. 1827. Included
also in Brown's Miscellaneous Botanical Works, Vol. I, 1866.
180

SUMMARY OF LITERATURE CITED 181

CARRUTHERS, W. On G-ymnospermatous Fruits from the Secondary

Rocks of Britain. Seemann's Jour. Bot. 5 : 1 seq. 1867.
CASPARY, R De Abietinearum floris feminei structura morpholo-

gica. Ann. Sci. Nat. IV. 14: 200-209. 1860.
CELAKOVSKY, L. Zur Gymnospermie der Coniferen. Flora 62 : 257-

264, 272-283. 1879.
Zur Kritik der Ansichten von der Fruchtschuppe der Abietineen.

Abhandl. d. konigl. bohm. Gesell. d. wiss. VI. 11: 1882.
Die Gymnospermen. Abhandl. d. konigl. bohm. Gesell. wiss. VII.

4: 1-148. 1890.
Die Gymnospermen ; eine Morphologisch-phylogenetische Studie.

Abhandl. d. konigl. bohm. Gesell. wiss. VIII. 4: 1890.
Nachtrag zu meiner Schrift iiber die Gymnospermen. Engler's

Bot. Jahrb. 24: 202-231. 1898.
Die Vermehrung der Sporangien von Ginkgo biloba L. Oester.

hot. Zeitsch. 50: 229-237, 276-283, 337-341. 1900.
CHAMBERLAIN, C. J. Winter Characters of certain Sporangia. Bot.

Gaz. 25: 125-128. pi 11. 1898.

Oogenesis in Pinus Laricio. Bot. Gaz. 27 : 268-280. pis. 4-6. 1899.
COKER, W. C. Observations on the Gametophyte and Embryo of

Taxodium distichum. Johns Hopkins Univ. Circ. 19: 45-46.

1900.
COSTANTIN and MOROT. Sur TOrigin des Faisceaux libero-ligneux

super numeral res dans la Tige des Cycadees. Bull. Soc. Bot.

France 32: 173. 1885. See review in Bot. Centralbl. 24: 101-

102. 1885.

COULTER, J. M. Notes on the Fertilization and Embryogeny of Coni-
fers. Bot. Gaz. 23: 40-43. pi. 6. 1897.
The Origin of Gymnosperms and the Seed Habit. Bot. Gaz. 26 :

153-168. 1898.
DE BARY, A. Comparative Anatomy of the Vegetative Organs of

the Phanerogams and Ferns. English translation. 1884.
DICKSON, A. Observations on some Bisexual Cones occurring in the

Spruce Fir (Abies excelsa). Trans. Edinburgh Bot. Soc. 6:

418-422. 1860.
Dixox, H. N. Fertilization of Pinus silvestris. Ann. Bot. 8 : 21-34.

pis. 3-5. 1894.
EICHLER, A. W. Ueber Welwitschia mirabilis, etc. Flora 47 : 459-

464, 473-479, 489-496, 508-510, 513-520. 1863.
Sind die Coniferen Gymnosperm oder nicht ? Flora 54 : 241-

247, 260-272. 1873 ; and later papers.

Coniferae in Engler and Prantl's Nat. Pflanzenfam. 21 : 108. 1889.
EXGELMANN, G. Morphology of the Carpellary Scales of Coniferae.

Am. Jour. Sci. III. 12: 469-470. 1876.
FANKHAUSER, J. Entwickelung des Stengels und Blattes von Ginkgo.

Bern. 1882.

182 MORPHOLOGY OF SPERMATOPHYTES

FUJII, K. On the Different Views hitherto proposed regarding the
Morphology of the Flowers of Ginkgo biloba. Bot. Mag. Tokyo
10: 15-25, 104-110. pi. 1. 1896.
On the Morphology of the Spermatozoid of Ginkgo biloba. Bot.

Mag. Tokyo 14: 260-266. pi. 7. 1900.
GOEBEL, K. Outlines of Classification and Special Morphology.

English translation. 316. 1887.

Organography of Plants. English translation. 153-155. 1900.
GOROSCHANKIN, J. Zur Kenntniss der Corpuscula bei den Gymno-

spermen. Bot. Zeit. 41 : 825-831. pi. 7 A. 1883.
Ueber den Befruchtung-Process bei Pinus Pumilio. 1883.
HIRASE, S. Etudes sur la Fecondation et I'Embryogenie du Ginkgo
biloba. Jour. Coll. Sci. Imp. Univ. Tokyo 8: 307-322. pis.
31-32. 1895.

Etudes sur la Fecondation et I'Embryogenie du Ginkgo biloba.
Jour. Coll. Sci. Imp. Univ. Tokyo 12: 103-149. pis. 7-9.
1898.
HOFMEISTER, W. Embryobildung der Phanerogamen. Jahrb. f. wiss.

Bot. 1 : 82-188. pis. 7-10. 1858.

Vergleichende Untersuchungen. English translation. 1862.
Lehre der Pflanzenzelle. 120. 1867.
HOOKER, J. D. On Welwitschia, a new Genus of Gnetaceae. Trans.

Linn. Soc. London 24: 1-48. pis. 1-14. 1863.
IKENO, S. Das Spermatozoid von Cycas revoluta. Bot. Mag. Tokyo

10: 367-368. 1896.

Vorlaufige Mittheilung iiber die Spermatozoiden bei Cycas revo-
luta. Bot. Centralbl. 69: 1-3. 1897.

Untersuchungen iiber die Entwickelung der Geschlechtsorgaiie
und der Vorgang der Befruchtung bei Cycas revoluta. Jahrb.
f. wiss. Bot. 32: 557-602. pis. 8-10. 1898.

JACCARD, P. Le Developpement du Pollen de VEphedra Helvetica.
Archives des sci. Phys. et Nat. III. 30: 280-282. 1893. See
review in Beih. zum Bot. Centralbl. 4: 230. 1894.
Eecherches embryologiques sur VEphedra Helvetica. Bull. Soc.
Vaudoise des sci. Nat. 30 : 46-84. pis. 3-10. 1894. See review
in Bot. Centralbl. 61 : 111-113. 1895.

JAGER, L. Beitrage zur Kenntniss der Endospermbildung und zur
Embryologie von Taxus baccata. Flora 86: 241-288. pis. 15-
19. 1899.

JURANYI, L. Bau und Entwickelung des Pollens bei Ceratozamia
longifolia. Miq. Jahrb. f. wiss. Bot. 8: 382-400. pis. 31-34.
1872.
KARSTEN, H. Beitrage zu Entwickeluiigsgesch. der Gattung Gnetum.

Bot. Zeit. 50 : 205-215, 221-231, 237-246. pis. 5-6. 1892.
Untersuchungen iiber die Gattung Gnetum. Ann. Jard. Bot.
Buitenzorg 11: 195-218. pis. 17-19. 1893.

SUMMARY OF LITERATURE CITED 133

KARSTEN, H. Zur Entwickelungsgesch. der Gattung Gnetum. Cohn's

Beitr. z. Biol. d. Pflzn. 6: 337-382. pis. 8-11. 1893.
LANG, W. H. Studies in the Development and Morphology of Cyca-

dean Sporangia. 1. The Microsporangia of Stangeria para-

doxa. Ann. Bot. 11 : 421-438. pi. 22. 1897.

Studies in the Development and Morphology of Cycadean Spo-
rangia. II. The Ovule of Stangeria paradoxa. Ann. Bot.

14: 281-306. pis. 17-18. 1900.
LOTSY, J. P. Contributions to the Life-history of the Genus Gnetum.

Ann. Jard. Bot. Buitenzorg II. 1 : 46-114. pis. 2-11. 1899.
METTENIUS, G. H. Beitrage zur Anatomic der Cycadeen. Abhandl.

d. konigl. Sachs. Gesell. d. wiss. 7: 565-608. pis. 1-5. 1861.
MIQUEL, F. A. W. Monographia Cycadearum. 1-82. pi. 8. 1841.
MOTTIER, D. M. On the Archegonium and Apical Growth of the

Stem in Tsuga Canadensis and Pinus silvestris. Bot. Gaz.

17: 141-143. pi. 8. 1892.

MURRILL, W. A. The Development of the Archegonium and Fertili-
zation in the Hemlock Spruce (Tsuga Canadensis). Ann. Bot.

14: 58a-607. pis. 31-32. 1900.
OERSTED, A. S. Bidrag til Naaletroeernes Morphologi. Videnskab.

Meddelelser Nat. Foren. Copenhagen. 1864.
PARLATORE, F. Studi Organografici sui Fiori e sui Frutti delle Coni-

fere. Opuscula botanica. 1864.
PAX, F. Allgemeine Morphologic der Pflanzen mit besonderer

Beriicksichtigung der Bliithenmorphologie. 1890.
REINKE, J. Parasitische Anabaena in Wurzeln der Cycadeen. Gott.

Nachr. 107. 1872.
SACHS, J. Lehrbuch. 1868.
SCHLEIDEN, M. J. Sur la Signification Morphologique du Placentaire.

Ann. Sci. Nat. Bot. II. 12: 373-376. 1839.
SCHNEIDER, A. Mutualistic Symbiosis of Algae and Bacteria with

Cycas revoluta. Bot. Gaz. 19: 25-32. pi. 3. 1894.
SCOTT, D. H. The Anatomical Characters presented by the Peduncles

of Cycadaceae. Ann. Bot. 11: 399-419. pis. 20-21. 1897.
Studies in Fossil Botany. London. 1900.
On the Occurrence of a Seedlike Fructification in certain Paleozoic

Lycopods. Ann. Bot. 14: 713-717. 1900.
SEWARD, A. C., and GOWAN, Miss J. The Maiden Hair Tree (Ginkgo

biloba L.). Ann. Bot. 14: 109-154. pis. 8-10. 1900.
SHAW, W. R Contribution to the Life-history of Sequoia. Bot. Gaz.

21 : 332-339. pi. 24. 1896.
SOKOLOWA, MLLE. C. Naissance de 1'Endosperme dans le Sac embry-

oniiaire de quelques Gymnospermes. Moscou. 1880.
SOLMS-LAUBACH, H. Die Sprossfolge der Stangeria und der iibrigen

Cycadeen. Bot. Zeit. 48: 177-187, 193-199, 209-215, 225-230. pi.

2. 1890.
13

184. MORPHOLOGY OF SPERMATOPHYTES

SOLMS-LAUBACH, H. Ueber die Fructification von Bennettites Gib-

sonianus Carr. Bot. Zeit. 48: 789-798, 805-815, 821-833, 843-847.

pis. 9-10. 1890.
Fossil Botany. Oxford. 1891.
STENZEL, G. Beobachtungen an durchwachsenen Fichtenzapfen. Ein

Beitrag zur Morphologic der Nadelholzer. Nov. Act. Nat. Cur.

38: 289-350. pis. 12-15. 1876.

STRASBURGER, E. Die Befruchtung bei den Coniferen. 1869.
Die Coniferen und die Gnetaceen. 1872.
Die Befruchtung und Zelltheilung. Jen. Zeitsch. 2: 435-536.

pis. 27-35. 1877.

Befruchtung und Zelltheilung. 1878.
Die Angiospermen und die Gymnospermen. 1879.
Neue Untersuchungen, etc. 1884.
Histologische Beitrage 3 : 151. 1891.
Ueber das Verhalten des Pollens und die Befruchtungsvorgange

bei den Gymnospermen. Hist. Beitr. 4 : 1-158. pis. 1-3. 1892.
TREUB, M. Recherches sur les Cycadees. Ann. Jard. Bot. Buitenzorg

2 : 32-53. pis. 1-7. 1881 (date of reprint, volume date being 1885).
Recherches sur les Cycadees. 3. Embryogenie du Cycas circi-

nalis. Ann. Jard. Bot. Buitenzorg 4 : 1-11. pis. 1-3. 1884.
VAN TIEGHEM, PH. Anatomic comparee de la Fleur femelle et du

Fruit des Cycadees, des Coniferes, et des Gnetacees. Ann. Sci.

Nat. Bot. V. 10: 269-304. pis. 13-16. 1869.
Recherches sur la Symetrie de Structure des Plantes vasculaires.

Ann. Sci. Nat. V. 13 : 1-314. pis. 3-8. 1870.
VON MOHL, HUGO. Ueber die Mannlichen Bliithen der Coniferen.

Verm. bot. Schriften 45-61. 1845. (Published as a dissertation,

1837.)
Ueber den Bau des Cycadeen-Stammes. Abhandl. d. konigl. d.

Acad. zu Miinchen 1: 1832. Republished and revised in Ver-

mischte Schrifteu 195-211. 1845.
Morphologische Betrachtung der Blatter von Sciadopitys. Bot.

Zeit. 29: 17-23. 1871.
WARMING, E. Recherches et Remarques sur les Cycadees. Oversigten

over d. K. D. Vidensk. Selsk. Forh. 1877.

Contributions a 1'Histoire naturelle des Cycadees. Ibid. 1879.
WEBBER, H. J. Peculiar Structures occurring in the Pollen Tube of

Zamia. Bot. Gaz. 23 : 453-459. pi. 40. 1897.
The Development of the Antherozoids of Zamia. Bot. Gaz. 24:

16-22. 1897.

Notes on the Fecundation of Zamia and the Pollen Tube Appara-
tus of Ginkgo. Bot. Gaz. 24: 225-235. pi. 10. 1897.
WIELAND, G. R. A Study of some American Fossil Cycads. Part I.

The Male Flower of Cycadeoidea. Am. Jour. Sci. IV. 7 : 223-

226. pis. 2-4. 1899.

iTURE CITED

185

WIELAND, G. R A Study of some American Fossil Cycads. Part III.

The Female Fructification of Cycadeoidea. Am. Jour. Sci. IV.

7: 383-391. pis. 8-10. 1899.
The Yale Collection of Fossil Cycads. Yale Sci. Month. 6: 1-11.

pi. 1. 1900.
WORSDELL, W. C. Anatomy of Stems of Macrozamia compared with

that of other Genera of Cycadeae. Ann. Bot. 10: 601-620. pis.

27-28. 1896.
The Anatomical Structure of Boivenia spectabilis. Ann. Bot.

14: 159-160. 1900.
The Structure of the Female " Flower " in Coniferae. Ann. Bot.

14: 39-82. 1900.
The Vascular Structure of the Ovule of Cephalotaxus. Ann. Bot.

14: 317-318. 1900.
The Affinities of the Mesozoic Fossil, Bennettites Gibsonianus.

Ann. Bot. 14: 717-721. 1900.
WOYCICKI, Z. On Fertilization in Coniferae (in Russian), pp. 1-57.

pis. 2. 1899.

APPENDIX

A PAPER by Mr. W. A. Murrill,* on Tsuga Canadensis, has
reached us too late to incorporate his results in the body of
the text.

The time which elapses between pollination and fertiliza-
tion is quite different from that we have given for Pinus. Two
weeks after pollination the archegonium initials become evi-
dent, one week later the neck of the archegonium is formed,
two weeks later the ventral canal cell is cut off, and five days
later fertilization occurs. As contrasted with Pinus, this period
of less than six weeks between pollination and fertilization is
remarkably short.

The neck of the archegonium exhibits considerable variation,
consisting of one to four cells. The ordinary number is two,
and the wall between them is usually oblique, although it may
vary in direction from transverse to longitudinal.

The author was not able to discover any passage of nu-
clei from the jacket cells to the central cell through wall
pores.

The two male cells are unequal in size, the diameter of
the functional one being twice that of the other. It would
be interesting to know whether this results from an unequal
division of the body cell, as in Taxus, or from unequal
growth.

The nuclei of both male cell and egg pass into the resting
stage before fusion, an observation directly at variance with
the reported behavior of these nuclei in Pinus.

* The Development of the Archegonium and Fertilization in the Hemlock
Spruce (Tsuga Canadensis). Ann. Bot. 14: 583-607. pis. 31-32. December,
1900.

187

188 MORPHOLOGY OF SPERMATOPHYTES

As in Pinus, the chromatin masses of the sex nuclei re-
main separate until segmentation into chromosomes takes
place. These chromosomes do not fuse, but mingle indis-
criminately near the center of the spindle. The evidence is
thus accumulating that male and female chromosomes may
remain distinct, and pass without fusion into succeeding cell
generations.

END OF PART I

BOOKS BY FRANK M. CHAPMAN.
BIRD STUDIES WITH A CAMERA.

With Introductory Chapters on the Outfit and Methods of the Bird
^Photographer. By FRANK M. CHAPMAN, Assistant Curator of Verte-
brate Zoology in the American Museum of Natural History ; Author of
"Handbook of Birds of Eastern North America" and "Bird-Life."
Illustrated with over 100 Photographs from Nature by the Author.
I2mo. Cloth, $1.75.

Bird students and photographers will find that this book possesses for them a unique interest and
value. It contains fascinating accounts of the habits of some of our common birds and descrip-
tions of the largest bird colonies existing in eastern North America; while its author's phenom-
enal success in photographing birds in Nature not only lends to the illustrations the charm of
realism, but makes the book a record of surprising achievements with the camera. The book
is practical as well as descriptive, and in the opening chapters the questions of camera, lens,
plates, blinds, decoys, and other pertinent matters are fully discussed, making the work an
admirable guide for the camera hunter.

BIRD-LIFE.

A Guide to the Study of our Common Birds. With 7 5 full-page Plates
and numerous Text Drawings by ERNEST SETON-THOMPSON. LIBRARY
EDITION, I zmo, cloth, $1.75; TEACHERS' EDITION, same as- Library Edi-
tion, but containing an Appendix, with new matter designed for the use of
teachers, and including lists of birds for each month of the year, izmo,
cloth, $2.00. Edition with 75 Colored Lithographic Plates, 8vo, cloth,
$5.00.

TEACHERS' MANUAL. To accompany Portfolios of COLORED PLATES of " Bird-
Life." Contains the same text as the Teachers' Edition of " Bird-Life, " but is
without the 75 uncolored plates. Sold only with the Portfolios, as follows :

PORTFOLIO No. I. — Permanent Residents and Winter Visitants. 32 plates.

PORTFOLIO No. II. — March and April Migrants. 24 plates.

PORTFOLIO No. III. — May Migrants, Types of Birds' Eggs, and 9 half-tone plates
showing Types of Birds' Nests. 34 plates.

Price of Portfolios, each, $1.25; with the MANUAL, $2.00. The three Portfolios,
with the MANUAL, $4.00.

HANDBOOK OF BIRDS OF EASTERN
NORTH AMERICA.

With Keys to the Species, Descriptions of their Plumages, Nests, etc., and
their Distribution and Migrations. With over 200 Illustrations. 1 2mo.
LIBRARY EDITION, $3.00; POCKET EDITION, flexible covers, $3.50.

D. APPLETON AND COMPANY, NEW YORK.

D. APPLETON & CO.'S PUBLICATIONS,
The Ice Age in North America,

AND ITS BEARINGS UPON THE ANTIQUITY OF MAN. By G.
FREDERICK WEIGHT, D. D., LL. D. With 148 Maps and Illustrations.
Third edition, containing Appendix on the "Probable Cause of Glacia-
tion," by WAEREN UPHAM, F. G. S. A., and Supplementary Notes. 8vo.
625 pages, and complete Index. Cloth, $5.00.

"Prof. Wright's work is great enough to be called monumental. There is not a page
that is not instructive and suggestive. It is sure to make a reputation abroad as well as at
home for its distinguished author, as one of the most active and intelligent of the living
students of natural science and the special department of glacial action." — Philadelphia
Bulletin.

The Great Ice Age,

AND ITS RELATION TO THE ANTIQUITY OF MAN. By JAMES
GEIKIE, F. R. S. E., of H. M. Geological Survey of Scotland. \\ ith Maps
and Illustrations. 12mo. Cloth, $2.50.

A systematic account of the Glacial epoch in England and Scotland, with special
reference to its changes of climate.

The Cause of an Ice Age.

BY SIR ROBERT BALL, LL. D., F. R. S., Royal Astronomer of Ireland,
author of " Starland." The first volume in the MODERN SCIENCE SERIES,
edited by Sir JOHN LUBBOCK. 12mo. Cloth, $1.00.

" An exceedingly bright and interesting discussion of some of the marvelous physical
revolutions of which our earth has been the scene. Of the various ages traced and located
by scientists, none is more interesting or can be more so than the Ice Age, and never have its
phenomena been more clearly and graphically described, or its causes more definitely located,
than in this thrillingly interesting volume." — Boston Traveller.

Town Geology.

By the Rev. CHARLES KINGSLEY, F. L. S., F. G. S., Canon of Chester.
12mo. Cloth, $1.50.

"I have tried rather to teach the method of geology than its facts ; to furnish the student
with a key to all geology ; rough indeed and rudimentary, but sure and sound enough, I
trust, to help him to unlock most geological problems which may meet him in any quarter of
the globe."— From the Preface.

An American Geological Railway Guide.

Giving the Geological Formation along the Railroads, with Altitude above
Tide-water, Notes on Interesting Places on the Routes, and a Description
of each of the Formations. By JAMES MACFARLANE, Ph. D., and more
than Seventy-five Geologists. Second edition, 1890. 426 pages. 8vo.
Cloth, $2.50.

" The idea is an original one, and its value is by no means confined to its originality.
The book does not confine itself to dry facts, but in a series of foot-notes gives a great deal
of information of general interest. The articles on the several States are by eminent scien-
tific authorities. Mr. Macfarlane has produced a very convenient and serviceable hand-book,
available alike to the practical geologist, to the student of that science, and to the intellient
traveler who would like to know the country through which he is passing."—
Evening Transcript.

New York: D. APPLETON & CO., 12 Fifth Avenue.

RETURN BIOLOGY LIBRARY

TO—* 3503 Life Sciences Bldg. 642-2531

LOAN PERIOD 1

ALL BOOKS MAY BE RECALLED AFTER 7 DAYS
Renewed books are subject to immediate recall

DUE AS STAMPED BELOW

HOI

Hi IN' fl 4'' 1986

••••..,.:•;

xJUil U * iouw
Subiect to Recall

Immediately

RQURNB)

• -ontsse

fiimnfiYIIRRARY

UNIVERSITY OF CALIFORNIA, BERKELEY
FORM NO. DD4 BERKELEY, CA 94720

LD 21-10m-7,'39(402s)

G.99C75

B