^>

IMAGE EVALUATION
TEST TARGET (MT-S)

/

//

/.

m

{'/ ^fj^

^. <

1.0 !!:■- I-

I.I

'^ IflM

li£ M

11.25 i 1.4

1.6

y.

^f.

^>

■'*j?

Photographic

Sciences
Corporation

23 WEST MAIN STREET

WEBSTER, N.Y. MS80

(716) 872-4503

rA%;:.v=." '-'^^.^^

!stair:aati*s^fe»cw*W!©5»Bfe:asa

^^mK,p!mmmm^m^m^^s^ismmmgssii^ifmBm>if^^!^Ki^ss.^^xsi^>smmmm^^^m^^SSMi

c^

s' 1

I

c

CIHM/ICMH

Microfiche

Series.

CIHM/ICMH
Collection de
microfiches.

Canadian Institute for Historical Microreproductions / Institut Canadian de microreproductions historiques

■^"m^xia&mmmWm^St^^^^^^^mfcm^^^^^mMmmstssii^

Technical and Bibliographic Notes/Notes techniques et bibliographiques

The Institute has attempted to obtain the best
original copy available for filming. Features of this
copy which may be bibliographically unique,
which may alter any of the images in the
reproduction, or which may significantly change
the usual method of filming, are checked below.

D
D
D

Coloured covers/
Couverture de couleur

Covers damaged/
Couverture endommagde

Covers restored and/or laminated/
Couverture restaurde et/ou pellicul6e

Cover title missing/

Le titre de couverture manque

I I Coloured maps/

Cartes gdographiques en couleur

□ Coloured ink (i.e. other than blue or black)/
Encre de couleur (i.e. autre que bleue ou noire)

□ Coloured plates and/or illustrations/
Planches et/ou illustrations en couleur

D
D

D

D

Bound with other material/
Reli6 avec d'autres documents

Tight binding may cause shadows or distortion
along interior margin/

La reliure serr^e peut causer de I'ombre ou de la
distortion le long de la marge int^rieure

Blank leaves added during restoration may
appear within the text. Whenever possible, these
have been omitted from filming/
II se peut que certaines pages blanches ajoutdes
lors d'une restauration apparaissent dans le texte,
mais, lorsque cela 6tait possible, ces pages n'ont
pas 6t6 film^es.

Additional comments:/
Commentaires suppl6mentaires;

L'Institut a microfilm^ le meilleur exemplaire
qu'il lui a 6t6 possible de se procurer. Les details
de cet exemplaire qui sont peut-dtre uniques du
point de vue bibliographique, qui peuvent modifier
une image reproduite, ou qui peuvent exiger une
modification dans la m^thode normale de filmage
sont indiquds ci-dessous.

I I Coloured pages/

D
D

Pages de couleur

Pages damaged/
Pages endommag6es

Pages restored and/oi

Pages restaurdes et/ou peiliculdes

Pages discoloured, stained or foxei
Pages ddcolordes, tachetdes ou piqudes

Pages detached/
Pages ddtach^es

Showthrough/
Transparence

Quality of prir

Quality indgale de I'impression

Includes supplementary materii
Comprend du materiel supplementaire

I — I Pages damaged/

I — I Pages restored and/or laminated/

I — I Pages discoloured, stained or foxed/

I I Pages detached/

I I Showthrough/

I I Quality of print varies/

I I Includes supplementary material/

Only edition available/
Seule Edition disponible

Pages wholly or partially obscured by errata
slips, tissues, etc., have been refilmed to
ensure the best possible image/
Les pages totalement ou partiellement
obscurcies par un feuillet d'errata, une pelure,
etc., ont 6t6 filmSes d nouveau de fapon d
obtenir la meilleure image possible.

This item is filmed at the reduction ratio checked below/

Ce document est film6 au taux de reduction indiqud ci-dessous.

18X 22X

10X

14X

12X

16X

20X

26X

30X

24X

28X

n

32X

\

re

6tails
>s du
nodifier
)r une
ilmage

es

The copy filmed here has been reproduced thanks
to the generosity of:

Library of Congress
Photoduplication Service

The Images appearing here are the best quality
possible considering the condition and legibility
of the original copy and in keeping with the
filming contract specifications.

Original copies in printed paper covers are filmed
beginning with the front cover and ending on
the last page with a printed or illustrated impres-
sion, or the back cover when appropriate. All
other original copies are filmed beginning on the
first page with a printed or illustrated impres-
sion, and ending on the last page with a printed
or illustrated impression.

The last recorded frame on each microfiche
shall contain the symbol -^ (meaning "CON-
TINUED"), or the symbol y (meaning "END"),
whichever applies.

Maps, plates, charts, etc., may be filmed at
different reduction ratios. Those too large to be
entirely included in one exposure are filmed
beginning in the upper left hand corner, left to
right and top to bottom, as many frames as
required. The following diagrams illustrate the
method:

L'exemplaire filmd fut reproduit grdce d la
gdn^rositd de:

Library of Congress
Photoduplication Service

Les images suivantes ont 6t6 reproduites avec le
plus grand soin, compte tenu de la condition et
de la nettet6 de l'exemplaire filmd, et en
conformity avec les conditions du contrat de
filmage.

Les exemplaires originaux dont la couverture en
papier est imprimde sont film^s en commen9ant
par le premier plat et en terminant soit par la
dernidre page qui comporte une empreinte
d'impression ou d'illustration, soit par le second
plat, selon le cas. Tous les autres exemplaires
originaux sont film^s en commenpant par la
premidre page qui comporte une empreinte
d'impression ou d'illustration et en terminant par
la dernidre page qui comporte une telle
empreinte.

Un des symboies suivants apparaftra sur la
dernidre image de cheque microfiche, selon le
cas: le symbole — ► signifie "A SUIVRE", le
symbole V signifie "FIN".

Les cartes, planches, tableaux, etc., peuvent dtre
filmds d des taux de reduction diffdrents.
Lorsque le document est trop grand pour dtre
reproduit en un seul clichd, il est filmd d partir
de I'angle supdrieur gauche, de gauche d droite,
et de haut en bas, en prenant le nombre
d'images ndcessaire. Les diagrammes suivants
illustrent la mdthode.

errata
i to

It

e pelure,

;on d

n

32X

\

1

2

3

1

2

3

4

5

6

i^fesaa-^-^^etffegfe^^'^?^^''^-^^"'

- Aarfj^ JMs^iiif^-

THE MORPHOLOGY OF THE CENTRAL CYLINDER
IN THE ANGIOSPERMS

By EDWARD C. JEFFREY, B.A. (Tor.), Ph.D. (Harv.)

'':>^

if

! I

I

l^-

, \

:;■::!

THE MORPHOLOGY OF THE CENTRAL CYLINDER
IN THE ANGIOSPERMS.

■ BY EDWARD C. JEFFREY. B.A., (TOR.), Ph.D. (HARV.)

V

i.jirt^'''^""*"^

-\5

-<

THE MORPHOLOGY OK THE CENTRAL CYLINDER
IN THE ANGIOSFERMS.

BY Edward C. Jeffrey, B.A. (Tor.), PhD. (Harv.)

(RepHnUd by permission f,om the Transactions of the Canadian Institute.)

The prevailing views in regard to the morphology of the vascular
apparatus of the higher Cryptogams and the Pha,-nogams are due to the
studies of Van Tieghem and Strasburger. Both these writers have
abandoned the conception of fibro-vascular strands, which is embodied in
the classic comparative anatomy of De Bary. In this treatise, the ax.al
organs, primary and secondary, of the higher plants are regarded as
composed of more or less modified fundamental tissue of a parenchy-
matous nature, bounded externally by the integumentary tissues of the
epidermis and traversed by conductive strands typically composed of
vessels and sieve-tubes. These fibro-vascular strands may consist of
clusters of xylem and phloem placed side by side, and in this case the
strands are called collateral fibro-vascular bundles. In another type the
phloem surrounds the xylem, and the bundle is then described as con-
centric. The latter type, by the localization of the phloem at intervals
on the outside of the xylem, becomes the radial bundle.

As a result of his study of the anatomy of the stem of the phaenoga-
mous order, Primulace^e, Van Tieghem' published in 1886 an import-
ant memoir on the morphology of fibro-vascu' ;■ strands, which has
formed the basis of most of the subsequent anatomic 1 work in this direc-
tion Although the general conclusions reached are founded on the study
of the Primulaces alone, Van Tieghem extends them to all vascular plants,
including the vascular Cryptogams. The primitive type of axis, accord-
ing to Van Tieghem, contains but a single concentric fibro-vascular
strand which is limited externally by a specialized layer, the pencycle.
The innermost stratum of the fundamental tissue, immediately adjoining
the pericycle, i^ also differentiated from the outer cortical tissue and is
named the endodermis. The fibro-vascular axis so delimited is caUed
the central cylinder or stele. This sort of fibro-vascular axis is found in
the stem of many Cryptogams and in the root of nearly all vascular

, Sur la Polyst^lie. Ann. de Sci. Nat. Bot., 7 •i^r., tom- 3-

'4

'1

It

1

4 JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANOIUSPERMS

plants, both cryptogamous and pha;nogamous. On account of the fact
that the stele or central cylinder in such axes is obviously single and
undivided, they arc designated inonostclic.

According to Van Ticghcm the monostclic central cylinder is some-
times modified by becoming expanded. The stele grows larger, and a
certain amount of the stelar parenchyma becomes aggregated in the
centre to form the medulla or pith. F"rom the pith, strands of paren-
chyma radiate outwards towards the pcricycle and constitute the medul-
lary rays. These parenchymatous rays divide the xylem and phloem
into segments, designated meristeles or bundles, which, in his opinion,
are morphologically very different from the concentric bundles of De
Bary. The pith and the medullary rays are also of a different morpho-
logical origin from the extracylindral fundamental tissue of the cortex.
The whole complex medullated central cylinder of this type is shut off
from the cortex by the same two layers, viz., the pericycle and the
endodermis, as is the primitive non-medullated concentric fibrc-vascular
strand. This modification of the central cylinder is found in the stems
of nearly all the Phanerogams, and, in isolated instances, among the
Cryptogams, e.g., the Osmundaceae, the genus Botrychium, and certain
species of Eqjisetum.

In the medullated monostelic axis, the endodermis and pericycle,
according to Van Tieghem, sometimes bend inwards between the meri-
steles and break into pieces at the bottom of the sinuosities. The frag-
ments of the thus interrupted pericycle and endodermis unite around the
inner side of the individual bundles, which they in the first place merely
subtended. The original medullated monostelic central cylinder is
considered to become, as a consequence, astelic, since its pith and rays
are imagined to become continuous with the fundamental tissue outside
the stele. The astelic type of central cylinder is found in certain amphi-
bious or limicolous phaenogamous orders and among the Cryptogams in
the genus Ophioglossum and certain species of Equisetum. In these
cases the separate meristeles, each surrounded by its own pericycle and
endodermis, may unite so that a more or less complete fibro-vascular
ring is formed, bounded both externally and internally by a continuous
pericycle and endodermis. The central cylinder is then said to be
gamodesmic. For example, in the genus Equisetum, E. limosum and
E. litorale have the individual meristeles completely surrounded by an
endodermis and pericycle ; in the rhizome of E. silvaticum and in the
aerial shoots of E. hiemale, on the other hand, the meristeles are fused to-
gether and the ring of united meristeles is bounded both outwardly and
inwardly by a circular endodermis. --

..?..! .

Vfc

-^i'iiWltW«'iaiii»ilriti.l-l*r.

^NGIUSPRRMS

)unt of the fact
jusly single and

^linder is some-
iws larger, and a
gregated in the
;rands of paren-
itute the mcdul-
1cm and phloem
I, in his opinion,
: bundles of De
iifferent morpho-
ue of the cortex,
type is shut off
ericycle and the
ric fibre-vascular
ind in the stems
nccs, among the
ium, and certain

is and pericycle,
etween the meri-
sities. The frag-
3 unite around the
first place merely
intral cylinder is
; its pith and rays
ntal tissue outside
I in certain amphi-
he Cryptogams in
isetum. In these
wn pericycle and
ilete fibro-vascular
^ by a continuous
then said to be
I, E. limosum and
surrounded by an
vaticiim and in the
steles are fused to-
oth outwardly and

JBFFRKV : MOR.MIOUKIV OK THE CENTRAL CVLINORR IN THE ANOIOMPERM8 5

Instead of expanding and becoming medullatcd monostclic, or pass-
ing from this modification int.. the astelic condition, the fibro-vascular
axi^ may, according to Van Tieghe.n, undergo successive bifurcations
and thus become polystelic. This type of central cylinder is very prcva-
lent among the vascular cryptogams, and occurs also in certain species
of I'rimula and Gunnera among the Phanerogams. In a cross section
of an older stem of this type, numerous concentric fibro-vascular strands
are to be seen imbedded in a matrix of fundamental tissue. Ihe origin-
ally separate steles of the polystelic axis, may unite, according to Van
Tieghem, to form a concentric annular stele, which is described as gamo-
stelic.

More recently, Van Tieghem'-* has returned to the subject of the central
cylinder of the Equisctace;e. In his essay on I'olystely, he de.scribes
the central cylinder of E. ar,>ense^ E. pratense, etc.. as niedullated
monostelic ; E. limosum and E. litorale as astelic, and the aerial shoot of
E hiemalc as gamodesmic. Confirming an earlier research of I fitzer
he recognizes that in the species originally described by him as medul-
lated monostelic, viz., £. .en'.«.. and £. />r«/^«^^. etc., there are, m the
region of the nodes, and at the basis of the smaller branches, well
marked indications of an internal endodcrmis, which disappears in the
internodes, only to recur in the successive nodes. He concludes that it
was primitively present throughout the entire length of the stem, and
has become in these species vestigial, persisting only at the nodes. He
expresses the opinion that the astelic type of central cylinder, as found
in E. limosum, is the primitive one, and that by fusion there resulted the
gamodesmic type found in the aerial shoots of E. hiemdlc and the
subterranean stem of E. silvaticum. Forms like E. arvense, E. pratense,
and E. scirpoides, are also gamodesmic, although the fact is obscured by
the partial degeneracy of the internal endodermis. As a result of this
investigation, it is apparent that the medullated monostelic type of
central cylinder does not exist among the Equisetacea;, and that those
species which at first sight appear to possess a stelar system of this type
are really degenerate astelic gamodesmic. The writer has recently
studied the development of the young stem in the Equisetacea^,* and has
shown that the stelar system in the young axis of this group, is
primitively gamodesmic, possessing a well-marked, continuous, internal
endodermis. The modification, in which the bundles are surrounded by
individual endodermal sheaths, appears quite late in the development of

TRemarkssurlaStruct. de la TigeclesPreies. Journ.de Hot., 4, '390. P- 365-

3 U. d. Schutzschfide d. Deutsch. Equiset., Jahrh. f. wiss. Bot. 6.

4 Mem. Bost. Soc. Nat. Hist. Vol. 5, No. 5, p. i7«-

\1

- ?
r

' , . .._„_, f'

JRPFRBV : MORPIIOl-Or.V Of TIIK CRNTRAI. CVLINnRR IN Tlllt ANOIOHPrRMS

the younj( sporophyte. If any (lci)endc«nce is to be placed on ontogenetic
results, it would appear that the primitive type of cauline stelar
arrari^'cment in the Ktiuisetaceie, is not one in which the bundles are
surrounded by individual endoderinal sheaths, but is, on the c(jnlrary,
that designated by Van 'ricghcm as gamodesmic. The appropriateness
of this term, as describing the conditions present in the young stelar
system of this group, will be discussed subsequently.

Van Tieghem' has recently re-examined the central cylinder of the
Ophioglossaceic, as a result of the discovery made by I'oirault" of the
occurrenccof an internal endodermisin the _>'tf«;/^'- stelar system of several
species of Hotrychiuin and Ophioglossum. I le confirms I'oirault's results
and describes the distribution of the endodermis in the young central
cylinder of /Mryc/u'um Ltotaria. The stelar system is tubular, and above
the point of origin of the first lcaf-tracehasaninternal,aswellasanextcrnal,
endodermis. The inner one rapidly disappears in the older region of the
young stem. A similar state of aflfairs occurs in Ophioglossum vuh^a-
turn, although in this case, the stelar tube becomes quickly broken into
separate strands on account of the (jverlapping of the foliar gaps. The
endodermis in this species also disappears in the older region of the
stem. I'oirault*' has described a somewhat persistent /«/^/'«n:/ endoder-
mis in O. Bcrgianum, O. capense, and O. ellipticum. Quite recently an
internal endodermal layer has been described by Farmer' as occurring
in the interesting genus Helminthostachys. His own observations, to-
gether with those of Poirault, lead Van Tieghem to the conclusion that
the cauline stelar system of the Ophioglossacea^ is, above the exit of the
first leaf-trace, astelic in the sense already defined. It is apparent also,
although Van Tieghem does not call attention to this fact, that the
primitive condition of the fibro- vascular tissues in the stem of this group,
is gamodesmic in his sen.se, and in this respect, corresponds with that
described by the writer as existing in the young stem of the Equisetacete.

The recent investigations of Van Tieghem and Poirault apparently
make it impossible to regard the central cylinder of the Equisetacea; and
Ophioglossaces: as in any sense, monostelic. Equisetum arvense, the
genus Botrychium and the genus Helminthostachys, which were de-
scribed by Van Tieghem in his essay on Polystely as belonging to the
latter type, are shown by a more complete study of their anatomy and
development to possess a gamodesmic central cylinder (in tlie sense of

5 Journ. de Bot., 1890, p. 405.

6 Anil. (1. Sci. Nat. Bot., 7 s^r., torn. 18., p. 169.
6a Op. Cit,, p. 169.
7 Ann. Bot.. vol. 13.. p. 434. ^

(VNniOKPRRMfl

J on ontofjciictic
cauline stelar
the bundles are
m tlie ccMilniry,
approijriutcness
Ke young stelar

cylinder of the
I'oirault" of the
system of several
I'oirault's results
e young central
bular, and above
ellasancxternal,
Jer region of the
'oglossum x'ti/'^a-
;kly broken into
iliar gaps. The
r region of the
nternnl cndoder-
uite recently an
er' as occurring
observations, to-
conclusion that
e the exit of the
s apparent also,
u's fact, that the
em of this group,
sponds with that
the Equisetaces.

irault apparently
Equisetacea; and
'iim arvense, the
which were de-
belonging to the
jir anatomy and
(in tiie sense of

JBPFRRY ! MORPHOLOGY OF THR CRNTRAI. CVLINnRR IN TIIR ANOIOJtPERMH 7

Van Ticghcm) in which the internal cndodermis has become more or less
obsolete. !t has further been rendered probable by the interesting in-
vestigations of I'oirault" that the apparently medullateil monostelic
central cylinder of the stem of the gleicheniaceous genus, IMatyzoma,
also possesses a pith derived from the extrastclar fundamental tissue.

The only remaining example cited by Van Ticghem in his classic
f essay on I'olystely, of a Cryptogam [jossessing a medullated rnono.stelic

central cylinder is that offered by the Osmundaceae. The writer hopes
to show in a memoir on the anatomy of the Fteridophyta, which will
appear .shortly, that here too, the pith is in reality an included portion of
the extrastelar fundamental tissue.

The medullated monostelic type of central cylinder may accordingly
be regarded as of very doubtful occurrence among the Cryptogams, and
In those cases where it is apparently present, it is derived from a modi-
fication of Van Tieghem's gamodesmic type. There would thus appear
to be very slight evidence for regarding the medullated central cylinder,
where it occurs among the Cryptogams, as derived from the dilatation
of an originally pithless stele. It is a well-established principle with
morphologists to attempt always the explanation of the structure of the
higher plants by the more easily understood corresponding features of
the lower groups. It seems to be in harmony with this method to eluci-
date the obscurities of the morphological interpretation of the cerlral
cylinder in the Angiosperms by the facts derived from the study of the
anatomy and development of the Pteridophyta and Gymnosperms.
The writer will shortly publish a memoir describing his studies on the
latter groups. In the present essay, he proposes to examine in a
general way, certain features of the structure and development of the
Angiosperms, which in the light of the investigations mentioned above,
seem to afford to some extent an elucidation of the morphology and
phylogeny of the higher Phienogams.

PROBLEMS.

The questions to be treated in the present memoir are briefly as follows :
Has Van Tieghem correctly described the mode of origin of his poly-
stelic type of central cylinder? Is his astelic type essentially different
from the polystelic ? Does the medullated monostelic central cylinder
ever arise by the dilatation of a primitively pithless stele ? What are the
salient anatomical features of the central cylinder of the Angiosperms ?

8 Op. Cit., p. 183.

™'WW^(&3HSfij(5%3

^f^€,'?;^«$¥;jBff6^^iffi«t!«iK

^^ur-f^asS***-**-

JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANGIOSPERMS

Do they throw any additional light on the difficult problems of the
genetic relationships of the group ? Anatomical studies have hitherto
been almost entirely neglected by American botanists on account of the
wealth of other matters, and that fact will serve as a sufficient justifica-
tion for the present research. Were any other needed, it would be
afforded by the extremely important phylogenetic results obtained in
recent years by English, French, and German palaeobotanists from the
study of the fossilized remains of the chiefly vegetative organs of various
groups of extinct Cryptogams. The advance of the science of Ecology
has furthermore made it less difficult to distinguish between those ceno-
genetic features of structure which are the result of the adaptation of
plants to their modern environment and those palingenetic traits which
serve as an indispensable guide in the interpretation of phylogeny.

THE POLYSTELIC TYPE.

In his essay on Polystely to which reference has already been made.
Van Tieghem' describes this modification of the central cylinder as origi-
nating from the pithless monostelic type, by the successive bifurcations
of the primitively simple stele. He has referred to it more recently in
practically identical terms."

In studying morphological problems, it is an accepted method of pro-
cedure to pass from the lower forms to the higher. This course has
been profitably pursued in the investigation of the morphology of
sporangia, anthers, ovules, etc., and in the examination of the homologies
of the gametophyte in the various groups of vascular plants. Curiously
enough, Van Tieghem has given very little attention to the lower forms
in his studies on the central cylinder, and still less to their development.
In his essay on Polystely he does not describe the development of any
cryptogamous stems of the polystelic type. It is only subsequently in
his Traits de Botanique (p. 765), that he makes a slight reference to the
development of the polystelic central cylinder of Pteris aquilina. Leclerc
du Sablon" has given an admirable description of the earlier stages in the
development of the stem of this species, but has apparently, not correctly
observed the later phases.

The writer proposes to describe briefly his own observations on this
form, as a preliminary to the examination of the phenomenon of

9 Op. Cit., p. 282.

10 Trait^ de Botanique, 1892, p. 1370 ; Elements de Botanique, 1898, p. 179.
II Ann. Sci. Nat. Bot., 7 s4r., 11 Tom., R^cherches sur la Tige des Fougires. ^

^gjfssagiBUtiiaHmm

E ANGIOSPERMS

: problems of the
ies have hitherto

on account of the
;ufficient justifica-
ded, it would be
jsults obtained in
Dotanists from the

organs of various
cience of Ecology
tween those ceno-
the adaptation of
;netic traits which
f phylogeny.

Iready been made,
,1 cylinder as origi-
sssive bifurcations
: more recently in

ed method of pro-
This course has
e morphology of
of the homologies
slants. Curiously
;o the lower forms
heir development,
velopment of any
y subsequently in
It reference to the
aquilina. Leclerc
irlier stages in the
ntly, not correctly

servations on this
phenomenon of

JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANGIOSPERMS 9

Polystely in the Angiosperms. This species is chosen because the
investigation of Leclerc du Sablon, and of the writer, show it to be quite
typical of the polystelic Pteridophyta. The form selected presents,
moreover, the phenomenon of Polystely in a high state of complexity,
and its cosmopolitan occurrence will make it easy for other botanists to
verify, if so minded, the accuracy of the description given here.

Leclerc du Sablon'' has correctly described the transitional region of
the young stem of P. aquilina. Higher up, and immediately above
where the first leaf-trace is given off from the concentric central cylinder,
the phloem sends a process into the centre of the vascular axis of the
strand. When the gap, caused in the axis by the exit of the foliar
tracheids, is closed again, the central phloem remains included, forming
a sort of pith. The exit of the traces of several subsequent leaves
causes similar gaps in the continuity of what is now the vascular tube,
and through these gaps the internal phloem communicates with that
outside Photograph i, plate 7, shows the structure of the young
stele, at a point where the second leaf-trace has just passed off. The
central island of small-celled bast can be distinguished, surrounded by a
ring of tracheids, which is interrupted opposite the leaf-trace, /. At the
level of the fourth or fifth leaf, the fundamental tissue penetrates to the
centre of the stele through the foliar gaps, forming, henceforth, a
continuous core within the internal bast. At this stage the stelar system
is a hollow cylinder, perforated by gaps above the points of origin of the
leal-traces. Leclere du Sablon" correctly describes it as being, in this
phase, like the stelar tube, which is permanently present in the stem of
Marsilea. The writer has not been able to distinguish any evidence of
the repeated bifurcation of the young central cylinder, described by Van
Tieghem as characteristic of his polystelic type. On the contrary, both
the writer's observations and those of Leclerc du Sablon seem to show
that the young stelar system of P. aquilina is of the type described by
Van Tieghem as gamostelic. That term, however, can hardly be
correctly used to describe the structure of the central cylinder of this
species, because it implies the fusion of steles, originally separate, of
which there is no indication whatever in this form. The writer"'' has
previously suggested the term siphonostelic as more accurately
describing the conformation of the young stelar system in the so-called
polystelic Filicales ; for his studies on the development of a number of

la Op. Cit., p. 3.
II Op. Cit., p. 5.

14 Trans. Brit. Ass. Adv. Sci., 1897, p. 869.

15 Mem. Bos. Soc. Nat. Hist. V. 5, No. 5, p. 160.

|''n'iii>Bmn- 11 III ininii Ti" ml ii nt -

-«*i3a«S!..-i3l©3."S«SL-^«Sii'«

■=*.3!rsS?RiBsfe**^5^>i-

JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANGIOSPERMS

representatives of this group have led to the conclusion that the young
central cylinder of this type is always tubular, and that there is no
indication of the bifurcation of the primitive fibro-vascular axis.

Photograph 2, plate 7, shows the older stelar tube of P. aciiilitia,
where it is provided with an axial core of fundamental tissue. At a is
to be seen the foliar gap corresponding to the leaf-trace /- ; r \s a root,
and /^ an earlier leaf When about a dozen leaves have been formed,
the vertical young stem of P. aquilina bifurcates, or in some cases
trifurcates, and the resulting divisions plunge into the soil and pursue a
subterranean horizontal course. Leclerc du Sablon'" has made a
curious mistake in regard to the mode of origin of the horizontal rhizome
of this form. He describes it as originating from a bud in the axil of one
of the leaves of the vertically-growing young plant. The writer has seen
the bifurcation of the stem regularly occurring in many hundreds of
examples of plants grown from different lots of spores and in different
years ; so that there can be no doubt that it is a quite normal process.
The writer's account, moreover, agrees exactly with the older description of
Hofmeister". Photograph 3, plate 7, represents a section of the young
stem just above the bifurcation of the stelar system ; a and b are the
tubular steles which are about to pass into the two horizontal
rhizomes. The stele a has just given off a leaf-trace/; in b the foliar gap
is also still open. Shortly after the young rhizomes make their
appearance, the leaves, which at first originate at small intervals, become
more widely separated, and their foliar lacunar frequently overlap. For
this reason, in a cross-section of the stem at this stage, one often sees an
appearance of independent dorsal and ventral steles, as in photograph 4,
plate 7. A series of sections, however, show that the stelar system is
still tubular. At this stage, a strand of brown sclerenchyma becomes
evident in the centre of the stelar tube, and, a few centimetres further
on, the ventral wall of the latter becomes involuted. A fibro-vascular
strand is subsequently detached dorsally from the involution, and forms
one of the two large axial concentric bundles found in the mature
rhizome. The single central strand is rapidly surrounded by a
sclerenchymatous tube formed from the sclerenchymatous rod described
above. This stage is represented in photograph 5, plate 7. Subsequently,
a second central strand is detached dorsally from the ventral wall of the
stelar tube. This strand is at first small, but ultimately becomes nearly
as large as the first-formed strand (fig. 6, plate 7). The two axial
concentric strands contribute to the formation of the leaf-traces, but
have no connection with the vascular supply of the roots.

16 Op. Cit., p. 5. ^

17 Higher Cryptog.inis, R.iy. Soc, p. 214.

^^-'-■ig^^-- ■ ) 1--^. — -

^"

ANGIOSPERMS

1 that the young
that there is no
lar axis.

; of P. aruilina,
tissue. At a is
:e /- ; r is a root,
ive been formed,
r in some cases
soil and pursue a
n'" has made a
)rizontal rhizome
in the axil of one
e writer has seen
any hundreds of
and in different
; normal process,
der description of
ion of the young
a and b are the
! two horizontal
in b the foliar gap
mes make their
intervals, become
tly overlap. For
one often sees an
in photograph 4,
stelar system is
nchyma becomes
ntimetres further
A fibro-vascular
lution, and forms
d in the mature
I r rounded by a
5us rod described
7. Subsequently,
entral wall of the
y becomes nearly
The two axial
e leaf-traces, but
ts.

JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANGIOSPERMS

II

Possibly misled by Hofmeister's older account, already referred to,
Van Tieghem"* has described the large axial strands as primitive, and
the strands outside as secondary cortical steles derived from these. A
study of development shows that, in reality, the external .strands are
primitive, as may indeed be inferred from the fact that the traces of the
roots and leaves are directly attached to them even in the mature
rhizome. The axial concentric strands, on the other hand, are of later
origin, and are to be regarded as medullary bundles.

It will probably be obvious to the reader who has followed the
foregoing account and examined the accompanying photographs, that
the development of P. aquilina offers little support to Van Tieghem's
hypothesis' of Polystely. The writer hopes to publish in the near
future, an account of the stelar development of a number of cryptogamic
forms of the so-called polystelic type, in which the arrangement of
the nascent fibro-vascular apparatus is identical with that found in
P. aquilina. The young stelar system of the so-called polystelic type
among the Filicales would appear to be characteristically tubular, and
the writer"' has already suggested on that account, that it may be
appropriately designated siphonostelic.

PRIMULACE.E.

Having discussed in a general way the development of the so-called
polystelic type in the Cryptogams, we may now profitably turn our
attention to the development of the same type in the Angiosperms. It
was the study of the anatomy of the stem of a large number of species
of the genus Primula which led Van Tieghem to propose the doctrines
in regard to the morphology of vascular strands, which are at present
so generally accepted by anatomists. Although derived, in the first
place, from an examination of certain Angiosperms, Van Tieghem's
hypotheses have been extended by their author also to the Cryptogams.

Van Tieghem describes two main types of cauline anatomy as
occurring in the Linnaean genus. Primula. In one type, the central
cylinder becomes dilated above the cotyledons and forms a medulla^"
This medullated monostelic central cylinder does not subsequently
undergo divisions. In a second type the central cylinder remains
undilated and without a medulla for several internodes above the

t3 Op. Cit., p. 765-

19 Trans. Brit. As^. Adv. Sci., iSq;, P- S69.

ao Sur. la Polyst^lie, p. 292-

i>IJii a ^JT'-

■^aa^aatai^aK^-a&^igimiieaa'^'i^^iai

*t4t^^tA^^^amim^!uiii^'^!h^^t*m

Liau U i*m*-i^^ mi ■ ■ Ji ittWiH futrtHu ilHUlff ,ir H lift! II li ■

12 JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANGIOSPERMS

cotyledons, and then bifurcates, repeatedly producing a varying number
of eccentric steles."' The fibro-vascuiar system in this case is
described as polystelic. Van Tieghcm retains, for species possessing a
cauline central cylinder of the first type, the original Linnaean generic
name Primula ; for the polystelic species, he revives Tournefort's genus
Auricula.

Van Tieghem does not seem to have followed closely the develop-
ment of the polystelic type in his genus Auricula. The writer has given
some attention to this subject, but it has not seemed necessary to make
his studies exhaustive, because an admirable account'"' of the whole
matter has recently appeared, with which the writer's own results are in
close agreement. Gwynne-Vaughan has examined the development of
the young stem and the relation of the leaf-traces to the cauline central
cylinder. He calls attention to the fact that the young pithless stele of
P. japonica and P. involucrata after the exit of from four to
eight leaf-traces becomes gamodesmic (gamomeristelic) in the same
manner as the central cylinder of the young stem of Botrychium
Lunaria, described by Van Tieghem. In other words, there is present
a collateral fibro-vascular cylinder with an internal endodermis as well
as an external one. This cylinder is perforated at intervals by the gaps
occurring at the points of exit of the leaf-traces, and around the margins
of these gaps the external and internal endodermis become continuous.
Higher up in the young stem, the gamodesmic cylinder becomes more
or less completely gamostelic by the formation of internal xylem and
phloem. The leaf-traces show the same variability as the cauline
strands. In the petioles of the younger leaves they are collateral
strands, while in the stalks of the later-formed foliar organs some of
them become concentric and would thus be considered, from Van
Tieghem's standpoint, as steles.

As regards the nature of Polystely itself, this writer reaches conclu-
sions which are so much in accord as far as they go with the investiga-
tions to be described in this memoir, that they may be quoted in full.
For example (p. 320) he makes the following statement : " Van Tieg-
hem seems to have entirely overlooked the all-important influence of
the leaf-traces on the phenomena of transition, and, indeed, on the vas-
cular system throughout the whole plant. On this account, he regards
Polystely, when present, as having originated by the continued bifurca-
tion of the central cylinder found in the lower part of the stem. He
speaks of it as flattening itself out, and constricting itself in the middle

21 Op. Cit., p. 305.

22 Gwynne-Vaughan ; Polystely and the Genus Primula, Ann. Bot., 1897. .^

ANGIOSPERMS

varying number
in this case is
cies possessing a
Linnaean generic
ournefort's genus

iely the develop-
2 writer has given
ecessary to make
It" of the whole
vvn results are in
; development of
le cauline central
I pithless stele of
•f from four to
ic) in the same
n of Botrychium
, there is present
dodermis as well
rvals by the gaps
3und the martjins
come continuous,
er becomes more
ernal xylem and
■ as the cauline
jy are collateral
organs some of
dered, from Van

reaches conclu-
'ith the investiga-
e quoted in full,
nt : " Van Tieg-
tant influence of
deed, on the vas-
ount, he regards
ontinued bifurca-
f the stem. He
self in the middle

JEFFREY: MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANGIOSPEKMS 13

until it becomes nipped into two. I have not seen anything that would
lead to a similar conclusion. The transitional phenomena in the seed-
ling and also the extreme variability in certain species of the most import-
ant characteristics of Polystely, give strong support to the opinion that
Polystely is not a primitive feature ot the group of Primulas, in which it
is found, but a comparatively recent modification." On a later page
(322) he writes: " One result appears clear: that in the Primulas the
gamostelic condition is more primitive and nearer normal monostcly
than is the dialystelic (polystelic) type, and that probably a gamodcsmic
condition preceded either."

The writer's observations have been made entirely on P. Auricula and
P. farinosa. The so-called monostelic species, P. sinensis, P. obconica,
and P. Forbesii, have been studied for comparison. The writer's results
in the case of P. Auricula correspond very closely to those reached in
the same species by Gwynne-Vaughan''^. As was long ago noticed by
Kamienski^ the epicotyledonary central cylinder remains an unmodified
single strand until several leaf-traces have been given off from it. About
the region of exit of the fourth or fifth leaf, it becomes a stelar tube with
the usual foliar gaps. As soon as the central cylinder becomes tubular,
the traces running to each leaf become three in number, the largest of
which comes off from the bottom of the foliar gap, while the two smaller
lateral traces are derived from the vascular tissues forming the sides of
the gap. It is only by following a series of sections that the really
cylindrical character of the stelar system can be made out, since the
foliar lacunae overlap and there is thus presented the appearance of com-
pletely isolated strands. It is only necessary in this connection to
emphasize the statements of Gwynne-Vaughan quoted above, that Van
Tieghem has overlooked the influence of the leaf-traces on the conforma-
tion of the central cylinder, and that there is no evidence of the repeated
bifurcation of the primitive stelar strand, such as he describes for this
and other so-called polystelic species.

The writer's study of the young stem of P. farinosa has resulted in
very similar conclusions. In this case the problem is simplified by the
fact that the individual leaves receive but a single strand from the cauline
stelar system. In the quite young axis, the stele does not immediately
become tubular. Above the point of origin of a leaf-trace, the pericycle
sinks into the stele, forming a sort of medulla (photograph 7, plate 8),
which may be compared with that composed of both phloem and peri-

S3 Op. Cit., p. 320.

24 Vcrgleich Anat. d. Prim., 1B78., p. lA.

-^.*&=i?a^^aafiriKinii6rf»*i>**^HBwwi:rrtHaic

a^i«5^ia4^afe.*wmi^r.frj^sae^!U^gs^^^£j.t;^^'4Earffei^»wair.^^(a»fca)W

11

*f:

-

14 JEFFREY ; MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANCIOSPERMS

cycle in the young fibro-vascular axis o{ Pteris aquilina (photograph I,
plate 7). After six or more leaf-traces have been given off, the stelar
tube encloses a core of fundamental tissue. The internal face of the
young stelar tube is at first devoid of phloem and it is only subsequently
that it appears, thus recalling the state of affairs described by Gwynne-
Vaughan in P. japonica and P. involiicrata. At this stage the stelar
tube would be described in accordance with Van Tieghem's terminology
as gamodesmic (gamomeristelic) : collateral strands occur even in the
old stem ; such a strand is figured by Kamienski."' Photograph 8,
plate 8, shows a transverse section of the stelar system of this species at
a region where the internal phloem has already made its appearance.
At r", a root is being given off; ^* is a radical stele which has already
made its exit from the central cylinder ; /' and /- are foliar gaps ; oppo-
site l> may be seen its corresponding leaf-trace. Photograph 9, plate 8,
is similarly lettered. In this case the leaf-traces corresponding to two
foliar gaps are to be seen. The stelar system of P. farinosa is thus
from the first, a tube, including primarily only a pericycle, and then
fundamental tissue as well. The tube is, in the beginning, collateral,
but becomes subsequently bicollateral through the appearance of inter-
nal phloem. The stelar tube has gaps in its walls above the points of
exit of leaf-traces. No such gaps occur opposite the outgoing radical
strands.

The writer's examination of the development of P. farinosa leads to
results similar to those obtained by Gwynne-Vaughan in the case of
other so-called polystelic species of this genus, viz., I. That the stelar
system in the young plant does not successively bifurcate, giving rise to
a varying number of steles, as described by Van Tieghem, but from the
first forms a stelar tube with foliar gaps. 2. That the stelar tube is
primitively collateral and only subsequently becomes bicollateral ; the
development of the internal phloem would seem to be a compensation
for the disappearance of secondary growth in the vascular system of
these peculiar species of Primula.

A

HALORHAGIDACEyE.

The peculiar conformation of the fibro-vascular system of the genus
Gunnera has been described by Reinke.'* Subsequently the central
cylinder of G. macrophylla has been somewhat exhaustively studied by

25 Op. Cit., plate 6, fig. 3.

96 Morpholcg., Abhand. v. Reinke, Leipzig, 1873.

utwwum-iyfflut^ww"*"— *"*

ANCIOSPERMS

a (photograph i,
:n off, the stelar
:ernal face of the
inly subsequently
)ed by Gwynne-
3 stage the stelar
em's terminology
occur even in the
Photograph 8,

of this species at
2 its appearance,
vhich has already
iliar gaps ; oppo-
)graph 9, plate 8,
esponding to two

farinosa is thus
icycle, and then
inning, collateral,
)earance of inter-
love the points of
: outgoing radical

farinosa leads to
lan in the case of
. That the stelar
ate, giving rise to
lem, but from the
the stelar tube is
bicoUateral ; the
)e a compensation
ascular system of

tern of the genus
ently the central
stively studied by

JEFFREY : MORPHOLOGY OF THE CKNTRAL CYI.INDF.R IN THE ANOIOSPK.RMS I5

Merker." Van Tieghem'-'" examined the anatomy of a number of
species of this genus in connection with his essay on Folystely.
Certain species. G. cordifolia, G. monoica, and G. prorepens possess mono-
stelic central cylinders which are quite pithless. In these species this
state of affairs is sometimes departed from in the leafy stems (as distin-
guished from the creeping rhizomes) where the stele may occasionally,
according to Van Tieghem,-" flatten out and bifurcate in the usual
polystelic fashion. The strands however in these cases are never more
than two. In G. magellanica and other species, the young stele bifur-
cates a certain number of times in the epicotyledonary region and
becomes polystelic. In species with large stems, e.s;., G. macrophylla,
etc., the bifurcations are described by Van Tieghem'" as being exceed-
ingly numerous.

Through the quite exceptional kindness of Dr. D. H. Scott, Director
of the Jodrell Laboratory, Kew, the writer has had the opportunity of
examining material of seedlings of Gunnera scabra. As Dr. Scott
himself has an article on this species in preparation, it will be possible
only to refer in a passing manner to the writer's own observations. All
that need be said in this connection is, that the epicotyledonary system
of G. scabra is primitively tubular, as are those of Pteris aquilina and
Primula Jarinosa ; and that it is further characterized by the same foliar
gaps subtending the points of leaf-traces. It resembles, moreover,
Primula japonica and P. farinosa, in the absence of internal phloem, on
the inner side of the young stele. So far as may be judged from the
development of these species, the polystelic Gunneras do not differ in
the nature of their young stelar system from the other polystelic forms
examined by Gwynne-Vaughan and the writer. They resemble the
polystelic Primulas in showing developmental evidence that the so-called
polystelic condition is derived from a primitively so-called astelic
arrangement of the vascular tissues.

SAXIFRAGACE^.

The writer has found the so-called phenomenon of Polystely to be
present in the Saxifragacese. As one of his students is working over
the anatomy and development of the stem in this order, it will be
necessary only to refer to certain features having an immediate bearing
on the present research.

27 Merker, Gunnera Macrophylla. Inang. Diss., 1888, Marburg.

28 Op. Cit., p. 307.

29 Op. Cit., p. 308.

30 Op. Cit., p. i<xt.

.^-f<»cy.,ij?w>a-j^v^.js3st4e»jwffvaft?^>i».a:a^-«flj^Mvfe!;.gt'a!j«Bg-Jg«te»*w"'»i.jwi ji. i

f

i6

JEFFREY : MORPHOLOGY OP THE CENTRAL CYLINDER IN THE ANGIOSPERMN

Ci

;>)

If a transverse section of the internodal region of Parnassia palustiis
be microscopically examined, it will be discovered that the stele is
provided with a pith formed of thick-vvalled collenchymatoid cells.
There is considerable secondary growth present in the wood" zone
traversed by slender medullary rays. The central cylinder issurroundcd by
a well-marked and lignified endodermis, which appears very clearly after
the use of phloroglucin and hydrochloric acid. Most of these features
may be seen in photograph 25, plate 11. In a section passing through
the point of origin of a leaf-trace, the side of a central cylinder is seen to
be hollowed opposite the outgoing trace ; at the bottom of the concavity
the phloem and xylcm are ab.sent, so that the collenchymatoid tissue of
the central cylinder of the stele is only separated from the external
fundamental tissue by the pericycle and endodermis. Not unfrequently
two leaf-traces come off from the stele, nearly opposite each other, as
may be noted in photograph 26, plate 11. When this happens the
central cylinder is split for a short distance into two strands. At their
lowest point the two strands are devoid of internal phloem and xylem, and
have along their inner borders a layer of the pith-like tissues described
above. The latter is separated from the fundamental tissue passing
through the stele of this region, only by the pericycle and endodermis.
These features may be seen in photograph 27, plate 11. Higher up, the
fibro- vascular tissues cover over the internal faces of the two strands
again, and they become, to employ Van Tieghem's terminology, two
separate steles, each of which apparently possesses a medulla of its own.
Photograph 26, plate 11, sufficiently illustrates the description given
above. The photograph very closely resembles Fig. 33, plate 15 of
Gunnera magellanica in Van Tieghem's memoir on Polystely.

The points of interest in the anatomy of this species seem to be that
there is present an apparently medullated monostelic axis, which
continues as such unless two leaf-trates from it come off close together ;
that under these circumstances the divided stelar axis becomes at first,
astelic, and then a little higher up, polystelic, in Van Tieghem's sense of
these terms. It is further to be noticed that both astely and polystely
are closely related to the exit of leaf-traces.

In Parnassta parviflora tne central cylinder is astelic, or rather gamo-
desmic, since it consists of a collateral fibro-vascular tube with an internal
endodermis, which communicates with that outside, through the gaps
occurring in the cylinder opposite the points of exit of the leaf-traces.
A section of the central cylinder of this species is shown in photograph
29, plate II: /is an outgoing leaf-trace ; g- is its corresponding gap,
and^^ is that of an earlier trace ; r and r^ are strands belon"ging to roots,

ANCIOSPERMS

rnassia palustris
;hat the stele is
:hymatoid cells.
ihe wood" zone

is surrounded by
very clearly after
3f these features

passing through
y'linder is seen to
\ of the concavity
rmatoid tissue of
om the external
^ot unfrequently
e each other, as
this happens the
rands. At their
-n and xylem, and

tissues described
tal tissue passing

and endodermis.
. Higher up, the
" the two strands
terminology, two
ledulla of its own.
description given
. 33. plate 15 of
)lystely.

; seem to be that
telic axis, which
aff close together ;
5 becomes at first,
rieghem's sense of
tely and polystely

c, or rather gamo-
be with an internal
through the gaps
of the leaf-traces,
wn in photograph
orresponding gap,
belonging to roots,

JEFFREY : MORPHOLOGY OP THK CENTRAL CYLINDER IN THE ANOIOSPERMS I?

which as is usual, cause no gaps in the fibro-vascular tissues, by their
exit from the stelar tube.

The foliar vascular strands in both these species are generally
throughout their cauline course collateral in structure, but early in their
petiolar course, they become concentric. In these features, they
resemble the leaf-traces oWrimulajaponica and P. denticulata described
by Gvvynne-Vaughan."

It is undesirable, for reasons already mentioned, to describe further
the anatomy of the Saxifragucc.x at the present time, and the subject of
Polystely may be left with a brief summary of the writer's conclusions.

From the study of the development of the so-called polystelic central
cylinder in the vascular Cryptogams and in the Angiosperms, the
writer has reached the following conclusions: (i) Polystely (so-called)
does not arise by the repeated bifurcation of the primitive epicotyledonary
central cylinder. (2) The young central cylinder of this type is primitively
tubular and is characterized by the occurrence of gaps in the wall of the
stelar tube corresponding to the leaf-traces. These essential features
may subsequently be obscured by the passage of numerous traces to the
same leaf, by the overlapping of the foliar gaps of different nodes, and
by the development of an internal or external system of strands (or
both) derived from the original stelar tube. (3) If the above conclu-
sions be accepted, the terms Polystely and Gamostely should apparently
be abandoned, since they involve an erroneous conception of the mode
• of development of the central cylinder in the various forms described
above. The writer proposes as more accurately describing the
nature of the central cylinder in this type, the term siphonostelic.'=

THE ASTELIC TYPE.

In his essay on Polystely, already often quoted. Van Tieghem
describes a modification in the central cylinder of vascular plants, which
he calls astelic. In this type, the epicotyledonary stele, having
previously become expanded and medullated, is said to become further
changed by the sinking in of the endodermis between the bundles. In
this case, according to Van Tieghem, the endodermis is apt to break at
the bottom of the undulations between the bundles, and the thus
separated segments then envelope the several vascular strands with

31 Op. Cit., p. 312.

32 Originally suggested in Trans. Brit. Ass. Adv. Sci., 1897, p. 869.

|8

JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANOIOSPERMS

individual endodcrmal sheaths. By this process the pith is supposed
to become continuous with the extrastelar fundamental tissue, and the
central cylinder disappears as such.

Van Ticghcm states that of closely allied species some may have a
medullated monostelic central cylinder, while others have the astclic
type of central cylinder. For example, in the Equisetaceae, E. arvense
is monostelic, E. limosum, on the other hand, is astelic. Among the
Phanerogams, Ranunculus aquatilis and R. Lingua are described as
having an astelic central cylinder, while a number of other species, e.g.,
R. acris and A', abortivus, are monostelic. In the genus Anemone, parallel
cases are presented by A. nemorosa, which is astelic, and A. pennsyl-
vanica which is monostelic. It has been pointed out in the introduction
that Van Tieghem in more recent publications has abandoned the idea
that any of the Ophioglossaceae or Equisetace.t are medullated mono-
stelic, but, on the contrary, regards them as astelic throughout the
epicotyledonary region of the stem. Although he has given up the
view that the cauline central cylinder of these Cryptogams is in any
case medullated monostelic, he is still of the opinion that certain
ranunculaccous genera" present examples of both the above-men-
tioned types of central cylinder.

Before going on to submit Van Tieghem's views in regard to the
morphology of vascular strands of certain Ranunculaceae to the test of
their agreement with the facts of anatomy and development, it is
necessary to describe certain modifications of these views which have
recently been proposed by Strasburger** and which have already been
accepted by many botanists of eminence. Strasburger admits the great
morphological value of Van Tieghem's definition of the central cylinder.
He suggests, however, inasmuch as the innermost layer of the cortex
which bounds the central cylinder is not always an endodermis,
i.e., a layer of cells provided with a cutinized undulating radial
band, but is very often composed of sclerenchymatous elements or
starch-bearing cells, that some general name is desirable for the internal
layer of the fundamental tissue adjoining the central cylinder. He
suggests the term phlceoterma as a suitable designation of the innermost
layer of the cortex, no matter what its histological character may be.
Strasburger questions the possibility of different species of the same
genus having in some cases a medullated monostelic central cylinder,
and in others presenting the astelic type of arrangement of their vascular
tissues. In other words, he believes it impossible that the medullary

33 £l^ment!^ de bot., 1898, p. 179.

34 Histolog. Beitrage III, pp. 309.313, 443, 443, 484.486.

ANGIOSPERMS

lith is supposed
1 tissue, and the

ame may have a
have the astclic
:acene, E. arvense
lie. Among the
ire described as
ither species, e.g.,
Knemone, parallel
and A.pennsyl-
the introduction
andoned the idea
ledullated mono-
; throughout the
las given up the
togams is in any
lion that certain
the above-men-

in regard to the
;eae to the test of
development, it is
views which have
have already been
:r admits the great
le central cylinder,
lyer of the cortex
s an endodermis,
undulating radial
latous elements or
ble for the internal
tral cylinder. He
m of the innermost
character may be.
pecies of the same
ic central cylinder,
jntof their vascular
that the medullary

JEFPRKV : MORPIIOLOOV OF TI.K CENTRAL IVI.INDER IN THE ANC.OSI'KRMS

I')

tissues should in some species of a given genus be intrastclar and in
otiicr species extrastelar. He accordingly regards the phUi^otcrma
which surrounds the whole complex of bundles in the so-called mono-
stelic type as morphologically different from the individual sheaths
enclosing the bundles in Van Tieghem's astelic type. Strasburger has
apparently overlooked the fact that Van Tieghem^^ had anticipated h.s
objection to the different morphological interpretation of the pith m
different species of the same genus, by announcing that m the case of
the cryptogamous orders, the Equisetacca; and Ophioglossace^, the
medullary tissue is always extrastelar, thus, as has already been pomted
out departing from the view originally expressed m h.s essay on
Polystely Van Tieghem'" still maintains, however, his earlier position
in regard to the varying morphology of the central cylinder in different
species of Ranunculus and Anemone.

We now come to the consideration of the development and structure
of so-called astelic axis in the Ranunculacea; and other groups.

RANUNCULACE^.

Our knowledge of ther anatomy, of this order is chiefly due to the
studies of Mari(^.^"' Although he has described quite exhaustively the
salient structural features of the root, stem, and leaves of this order, he
has "iven little or no attention to the subject of the development of the
stem" which, as the writer hopes to show, is of considerable importance
from' a morphological standpoint. Quite recently Jancezewsk,- has
examined the anatomy of a number of species of Anemone. Stras-
burger'" attributes to Mari6 the statement that Ranunculus acrts pos-
sesses a common endodermis (phloeoterma). i.e., in other words, that the
central cylinder of this species is meduUated monostelic. Manes
description*' is extremely condensed, and, on that account, is some-
what ambiguous ; but the writer is inclined to interpret it as mean-
ing that the bundles of this species have individual endodermal sheaths
since he states that R. acris resembles R. multifidus, which has individual
endodermal sheaths.

35 Journ. de Bot., 1890.

36 Elements de Botanique, p. 179.

37 R^cherches sur la Structure des R<Snonculacee8. An... de sci. Nat, Bot., 6 s6r., tom. .0.

38 R^vue G^n^rale de Botaniqu.. , .898. fetudes Morphol. sur le Genre Anemone.

39 Op. Cit., p. 3>''

40 Op. Cit., p. 80.

20 JRPrRRV ; MORPIIOI.Or.Y or TIIR CINTRAL CVLINnRR IN THR AN(]|OIPBRM8

The yoiinfj stem of A', acn's has a stelar system very much like that of
Botrychium Lunaria, i.e., it consists of a fibro-vascuiar tube pierced by
foliar gaps. Through the latter, the internal phlceoterina, which is ex-
tremely well marked in the young stem of this species, communicates
with the external phhcoterma. Photograph 13, plate 9, shows a leaf-
trace passing off" from the stcliir tube, and subtending it is the foliar gap.
The continuity of the internal anfl external phheotermal layers through
the foliar gap can be easily distinguished. The photograph in this case
was made from a section treated with phloroglucin and hydrochloric acidi
which causes the lignified phkcoterma to stand out sharply. The stelar
tube, as in the case of the young so-called polystelic axes, only becomes
fully established after several leaf-traces have been given off, including
those of the cotyletlons, and at first, as in the young stem ot Primula
farinosa, the tube contains only enclosed pericycle and no fundamental
tissue. The leaf-traces to each leaf arc originally single, but very soon
become three in number : a central large one and two smaller lateral ones.
The departure of the median trace causes a large gap in the stelar tube,
through which, the internal and external phlceotermal layers communi-
cate. The lateral strands cause smaller breaks in the continuity of the
fibro-vascular cylinder and frequently the internal and external cortex do
not become continuous through them. The gaps, as a result, are occu-
pied merely by pericycle. The stelar cylinder in the more advanced
young stem is characterized by the overlapping of the foliar gaps, so that
in a cross section it appears as a circle of separate bimdles. This state
of affairs is shown in photograph 14, plate 9. Somewhat higher up in
the young stem, the overlapping foliar gaps become more numerous,
both on account of the increased number of leaf-traces given off" to each
leaf and the greater elongation of the individual foliar gaps. For
this reason, the stelar tube, in transverse section presents the appearance
of a number of separate vascular strands, the fibro-vascular bundles, each
of which is surrounded by its own phlceotermal sheath. In older regions
of the stem, the fibro-vascular strands of the leaf-traces dip inwards and
pass outwards again, thus simulating the peculiar course of the leaf-
traces which becomes the rule among the Monocotyledons.

From the thick parenchymatous hypogjeous stem of R. acri.<i the
more slender aerial shoots originate. Since the latter must support, in
spite of their comparative delicacy, a considerable weight of leaves,
flowers, and fruit, the mechanical tissues which are scarcely present at
all in the subterranean stem, are well developed in the epigsous shoots.
As is well known, the mechanical tissues of R. acris appear as fibrous
sheaths around the individual bundles. Already, in the older subter-

I

luch like that of
ubc pierced by
na, which is ex-
i, communicates
), shows a leaf-
s the foliar ^'ap.
1 layers through
raph in this case
ydrochloric acidi
ply. The stclar
2S, only becomes
-en off, including
tem ol Primula
no fundamental
Ic, but very soon
aller lateral ones.
1 the stelar tube,
layers communi-
continuity of the
eternal cortex do
result, arc occu-
more advanced
bliar gaps, so that
idles. This state
vhat higher up in
more numerous,
given off to each
oliar gaps. For
its the appearance
ular bundles, each
, In older regions
s dip inwards and
jurse of the leaf-
dons.

1 of R. acrts the
r must support, in
weight of leaves,
scarcely present at
e epigajous shoots,
r appear as fibrous
the older subter-

,KKrR«V . MORHH0U.GV Of THK CKNTRAL CVUINDKR IN TIU ANGIOHPFRM^

21

rancan stem, the internal phluH,terma is less apparent ana .a the base
of the aerial sh<.otH, both external and i.Uernal phlcL-otcrmal layers
cease to be recognizable as such This disappearance takes place
pari Mssu with t. -^ vppeafftncc of the srlerified sheaths ab"ut the
individual bundles. i" is highly probable that the ""^cr ayer of
sclerified cells on the extcfM.l and internal faces of the bundle-sheath
represents the phl<«oterma. and that the rest of the fibrous t.ssue .s to
be regarded as sclerified pericycle. It is not easy to come to an op.n.on
In regard to the position of the phheoterma on the flanks of -he
bundles, but since the larger so-called bundles arc completely sur-
rounded by a phlaeotermal sheath in the hypoga,-ous stem, it may be
safely assumed that the larger bundles of the aerial shoots arc to be
considered as morphologically bounded by a similar membrane The
smaller bundles in the hypog^eous stem are often united with he
larger ones by their pericycle; as a consequence, they arc surrounded
by the same phlceotermal sheath. It .seems probable that they are to
be regarded as similarly united with the larger bundles in the ep.ga.ous
shoots.

The development of R. sceleratus has been studied by the writer, and
does not differ essentially from that described in R. arm, in respect to
the structure of the young central cylinder and the di.stnbut.on of the
phlceotermal sheath.

The young stem of R. repens also resembles closely that of A', acris
in every respect, both as regards development and the distribution of
phloeoterma. In the older stem of this species, the phlceotermal sheath
disappears completely in the internodes of its characteristic runners ; .t
becomes distinct again, however, in the rooting nodes, both internally
and externally.

R abortivus is described by Marid*' as having an external phloeo-
terma The development of this species is, accordingly, of special
interest The conformation of the young fibro-vasc ilar tube does not
differ in any important feature from that of R. acris, and consequently
need not be described. The distribution of the phlceoterma in the
youn- stem of this species is, however, of considerable morphological
interest. In the lower region of the epicotyledonary central cylinder,
an internal phlceoterma is present, which communicates with the
phlceotermal layer outside the fibro-vascular tube, through the foliar
gaps, in a manner quite similar to that obtaining in R. acrts, as
described above. The internal phlceoterma is somewhat persistent,

4X Op. Cit., p. 8a.

^..^ ■•;H!fc--'J wr^sf ■ -

.X.

-.' ^^z*.''r^v',t^^

22 JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANGIOSPERMS

and only in rather stout younj^ hypogaeous steins does it become quite
obsolete.

Photograph 15, plate 9, represents a section through the upper region
of the stem of a well-advanced seedling. Phloroglucin and hydro-
chloric acid have been used to bring out the lignified phloeoterma
which can be clearly distinguished on the outer margin of the fibro-
vascular segments, but less distinctly along their inner border. The
occurrence of an internal phloeoterma in the young fibro-vascular cylinder
of a species in which it is absent in the mature hypogjeous stem, is of
considerable morphological interest. In the aerial shoots, the phloeo-
termal boundaries are no longer distinguishable histologically, although
on morphological grounds they should doubtless be regarded as never-
theless present.

Even in the quite young stem of R. rhomboideus, there is no trace of
an internal phlojoterma. However, by examining the youngest region
of the axis, a quite distinct internal phloeoterma may be demonstrated.
Photograph 17, plate o, represents a cross section of the young stem
where the fibro-vascular cylinder appears as an unbroken ring. Photo-
graph 18, plate 9, shows a tranverse section of the stelar tube where it is
interrupted by a foliar gap ; the junction of the internal and external
phloeotermal layer around the sides of the foliar gap can be clearly
discerned. Photograph 16, plate 9, represents a section through an
older region of the subterranean stem ; three foliar gaps are to be seen
and the internal phloeoterma has now become quite obsolete. The
aerial shoots of this species do not differ materially in the structure of
their vascular strands from the other species already described. R.
rhomboidetis is distinguished from all the species examined by the com-
plete disappearance of any histological evidence of the existence of an
internal phloeoterma, even in quite young stems. Even here, the
boundary of the cortical tissues on the inside of the stelar tube may
readily be determined by examining the stem in a very early stage of
development.

The writer has not been able to secure seedlings of other species of
Ranunculus, but those described appear to be sufficiently varied in their
affinities to afford reliable data for conclusions as to the correct morpho-
logical interpretation of the central cylinder of this genus. These
deductions may, however, be advantageously deterred until the allied
genus Anemone, has been considered.

As has been already mentioned, Van Tieghem states that certain
species of Anemone have a medullated monostelic central cylinder

JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANG.OSPF.RMS

23

; ANGIOSPERMS

s it become quite

. the upper region
lucin and hydro-
ified phlceoterma
rgin of the fibro-
ner border. The
(-vascular cylinder
ogaeous stem, is of
hoots, the phloeo-
ogically, although
egarded as never-

liere is no trace of
e youngest region
be demonstrated.
f the young stem
3ken ring. Photo-
lar tube where it is
:rnal and external
gap can be clearly
ection through an
raps are to be seen
ite obsolete. The
in the structure of
ady described. R.
mined by the com-
the existence of an
Even here, the
the stelar tube may
very early stage of

of other species of
ently varied in their
the correct morpho-
this genus. These
;rred until the allied

. states that certain
lie central cylinder

while other species are astelic. A nemonc nemorosa belongs to the latter
class and is described by Mari^- as having the individual bundles sur-
rounded by phlceotermal sheaths. A. pennsylvanica has, accordmg to the
same author," a continuous external phlceoterma surroundmg the
whole complex of bundles. The writer's discovery of an mternal
phloeoterma in the young stems of various so-called monostehc speces
of Ranunculus, has suggested the examination of the younger region of
the subterranean axis of this species of Anemone. Photograph 19.
plate 10, reproduces a section of the quite young hypoga;ous stem ot
A pennsylvanica. The stelar system is interrupted by a foliar gap.
around the margins of which the internal and external phlceotermal layers
are in communication. Photograph 20, plate 10, represents an older
region of the cylinder where it is interrupted by numerous foliar gaps,
The segments of the stelar tube are, in this case, individually sur-
rounded by phlceotermal sheaths. In the older region of the hypog^-
ous stem, it becomes continually more difficult to distinguish he
phlceoterma on account of the sclerification of the pencycle, and, ulti-
mately, of the phloeoterma itself.

The young stem of ^. cylindrica has also been studied in this connec-
tion but in this species, even the outer phlceoterma is very poorly
devdoped, and the internal one is obsolete throughout. The absence o
a well-marked phloeoterma in the hypogaeous stems of the genus
Anemone has been especially noted by Janczewski.^^and accordingly,
it is not surprising that this feature should be exemplified by A. cyUn-
dtica.

The writer has also examined the stem of the closely allied genus
Hepatica. In the seeding of Hepatica triloba, there is a well-marked
external phloeoterma, but the internal phlceotermal sheath is not differ-
entiated at all on account of the sclerification of both the pencycle
and the external layer of the medulla. The latter features may be
seen in photograph 30, plate 11.

In connection with the Ranunculaceae may be described the structure
of the young stelar system of Sarracenia purpurea. In this speces, the
young central cylinder is tubular with foliar lacun.. The externa
phloeoterma is poorly developed and the corresponding internal layer
cannot be distinguished even in the very young axis.

The reader who has followed the descriptions given in the foregoing

42 Op. Cit., p. 56.

43 0p. Cit., p. 62.

44 Op. Cit., p. 3-

mM

24 JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANGIOSPERMS

paragraphs of the development of the stelar system of Ranunculus and
Anemone, will doubtless be struck with the bearing of the facts there
described, on the problem of the morphology of the central cylinder
in the Angiosperms. For example, we find in this order that the astelic
type of Van Teighem does not originate by the sinking in of the
phloeoterma around the individual bundles, with a subsequent union of
the individual phlceotermal sheaths to form a continuous external and
internal phloeoterma, because the fibro-vascular system is, from the very
first, tubular, and the central cylinder has primitively an internal
phloeoterma. Further, the stelar system of so-called monostelic species
of Ranunculus and Anemone appears from a study of the coarse of
development to be really an astelic type, in which the internal phloeoterma
has become obsolete. These two facts, viz., the primitively tubular
(gamodesmic) nature of the central cylinder in the Ranunculaceae, and the
derivation of the so-called monostelic medullated type by the degenera-
tion of the internal phloeoterma, are, in the writer's opinion, of great
importance in connection with the view to be taken of the morphology
of the central cylinder in the Angiosperms generally. Before proceeding
to consider that subject, however, it will be well to describe the develop-
ment of representatives of other angiospermous groups, which present
the phenomenon of astely, so-called.

NYMPH^ACE/E.

In his essay on Polystely, Van Tieghem describes the Nymphseaces as
an order illustrating his astelic type of central cylinder. The
arrangement of the fibro-vascular strands in the mature stem of the
various genera of this order is extremely complex, and, on that account,
it is the more desirable to investigate the development of the stelar
system in the young plant in order to discover, if possible, what is really
the primitive condition of the fibro-vascular apparatus in this group.

The genus Brasenia is probably one of the most primitive of
the order. In the internodes of the rhizome of the mature plant
of Brasenia purpurea, there are present two concentric fibro-vascular
strands, in which the elements of the xylem, as is commonly the case in
aquatics, are represented merely by air-spaces. The epicotyledonary
stele of the seedling of Brasenia purpurea is a pithless fibro-vascular
strand, in which the vessels are not degenerate as in the mature rhizome.
The stelar system retains the simple character until a considerable
number of leaf-traces have been given off. The stem, however, sooner
or later becomes relatively massive, and the fibro-vasGular system

ai*^-

ANGIOSPERMS

Ranunculus and
f the facts there
; central cylinder
er that the astelic
nking in of the
sequent union of
3US external and

is, from the very
vely an internal
nonostelic species

of the coarse of
ernal phlceoterma
rimitively tubular
inculaceae, and the

by the degenera-

opinion, of great
f the morphology
Before proceeding
cribe the develop-
ips, which present

e Nymphaeaceae as
cylinder. The
ature stem of the
i, on that account,
-nent of the stelar
sible, what is really
i in this group.

nost primitive of
the mature plant
itric fibro-vascular
imonly the case in
le epicotyledonary
iless fibro-vascular
le mature rhizome,
itil a considerable
m, however, sooner
o-vasGular system

JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANGIOSPERMS 25

becomes tubular, as in the young stems of the ranunculaceous species
described above. The tube has, likewise, both an internal and external
phlceoterma, and the stelar tube is characterized by the same foliar gaps,
around the margins of which the internal and external phlceotermas
communicate. When the young stem has reached a length of from four
to eight centimetres, the foliar gaps begin to overlap, and the two
concentric strands which are at present in the mature stem make their
appearance. When the stem passes out of the mud into the water
to form the well-known floating shoots, the xylem degenerates and is
represented merely by a cavity in each of the two stelar strands.
Photograph 22, plate 10, represents a section through the young tubular
stelar system of Brasenia purpurea, at a point where a leaf-trace is
being given off, and is subtended by its corresponding foliar gap. It
will be obvious from the above description and the accompanying
photograph that the young stelar system of Brasenia purpurea is
primitively tubular.

Two species of Nuphar have also been examined in this connection.
Nuphar luteum, var. Kalmianum, is very favorable for study on
account of the relative simplicity of its stelar system. The young
epicotyledonary central cylinder is here also, at first, a pithless strand,
and only after the exit of several leaf-traces does it become tubular with
a pith of fundamental tissue. The stelar tube is provided with an
internal as well as an external phlceoterma. Photograph 22, plate lO,
represents a section through the young rhizome at a point where a leaf-
trace has just been given off and a second one is in the act of making
its exit from the tubular central cylinder. In the older rhizome, the
foliar gaps begin to overlap, so that in a transverse section the stelar
tube seems to be broken up into two or more segments, each surrounded
by its own phloeotermal sheath. The number of fibro-vascular segments
become subsequently increased by the fact that .-everal leaf-traces are
given off to each leaf. Finally the stelar tube becomes entirely
unrecognizable by reason of the increasing complexity in the arrange-
ment of the vascular strands.

The stelar development of Nuphar advena is very similar to that of
the species just described. In this case, the epicotyledonary fibro-
vascular strand passes quite rapidly into the tubular condition. The
fibro-vascular cylinder has likewise internal and external phlceotermas
which communicate through the foliar gaps. Very scon the leaf-traces
to the individual leaves become numerous, and the stelar tube is
consequently interrupted by a large number of foliar gaps. The com-
plexity is subsequently further increased by the appearance of an

26 JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER .N THE ANG.OSPERMS

internal and external series of strands, and the stelar tube can no
longe. be recognized as such. Photograph 23. pUte 10. represents a
section of the young stem of Nuphar advena in a region where the
leaves have numerous traces and the stelar tube, as a consequence, .s
broken into many segments.

The young stelar system of NymphcBa zanzibarensis is at first a
pithless strand, but very rapidly becomes tubular with usual foliar gaps
Quite early all semblance of the original stelar tube vanishes on accoun
of the appearance of complexly anastomosing internal and external
systems of fibro-vascular strands.

The young stem of Nymphm tuberosa has also been studied by the
writer The epicotyledonary central cylinder is originally pithless, as
ZL other TympLaceous species already described, but rapidly
expands and almost immediately breaks up into an ^^''^^^\'^':'^^^^
system of anastomosing bundles, so that it is extremely difficult to
recognize the existence of a stelar tube. Towards the end of a season .
growth, however, the stelar tube again becomes obvious in the more
slender region of the stem which is formed during the autumn. Ihe
stelar tub^ in this case has an internal phloem as well as an internal
phloeoterma. The tubular condition is probably to be regarded
as a reversion to the primitive type of stelar system, and may be
compared with that noticed by Gwynne-Vaughan^^ m the autumnal
region of the stem of Primula obtusifolia and P. tnvolucrata. In
another essay this writer^e has described the occurrence of concentric
fibro-vascular strands in slender pedicles of the tubers which are so
characteristic a method of vegetative reproduction m this species l
have examined the structure of the pedicles m -^--^^^fj^^l
confirmed the existence of the concentric strands described by Gwy nne
Vaughan. The fibro-vascular tissues in this case form a tube with foliar
gaps a fact which seems to have escaped his notice. As the slender
pedide passes into the tuber, the tubular concentric stele vamshes in an
extremely complex system of anastomosing strands, some of which are
concentric while others are merely collateral The -appearance of
the tubular stelar system, in the autumnal region of the young
rhizome and in the pedicles of the tubers is interesting as a Probable
case of reversion. The base of the lateral shoots of the rhizome of
Fteris aquilina not unfrequently shows a reversion to the more prim.trve
tubular stele with a single medullary strand. The writer has found a
sLilar concentric tubular stelar system in the pedicles of Nymph<Ba
dentata. ^

45. Op. Cit., p. 309-

46. Ann. Bot., v. lo, p. 290.

E ANGIOSPERMS

:lar tube can no

e lo, represents a

region where the

a consequence, is

nsis is at first a
1 usual foliar gaps,
anishes on account
rnal and external

jen studied by the
ginally pithless, as
ribed, but rapidly
3xtremely complex
:remely difficult to
\e end of a season's
jvious in the more
the autumn. The
well as an internal
f to be regarded
stem, and may be
I*'* in the autumnal
P. invelucrata. In
•rence of concentric
;ubers which are so
I in this species. I
question and have
;scribed by Gwynne-
irm a tube with foliar
:ice. As the slender
; stele vanishes in an
s, some of which are
rhe reappearance of
;gion of the young
resting as a probable
s of the rhizome of
to the more primitive
he writer has found a
jedicles of Nymphaa

JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE

ANGIOSPERMS

27

Through the kindness of Mr. Jackson Dawson, of the Arnold Arbore-
tum of Harvard University, the writer has had the opportunity of study-
ing the anatomy of a number of seedlings of Nelumbium ititeum. The
first root in this species is abortive, so that in a series of sections through
the base of the young plant, one passes almost immediately into the coty-
ledonary region. The very short, pithless central cylinder of the young
axis becomes tubular immediately below the level of exit of the coty-
ledonary traces. The exit of the latter breaks the stelar tube into four
segments. Photograph 10, plate 8, shows the disposition of the stelar
system just above the point where the cotyledonary traces have passed
off. One cotyledon, cot., is present; the other has been broken off. The
four cotyledonary traces anastomose outside the lacunar cortex and the
resulting network gives off the cotyledonary strands proper, as well as a
series of strands which run through the cortex of the first internode.
Photograph 24, plate 10, rep-esents a section through the first internode ;
a are the four primitive fibro-vascular strands ;/are the cortical strands.
At the second node, the four central strands again form a tube from
which a number of functional roots are derived, as well as strands for the
third leaf The exit of the latter causes a breaking up of the stelar
tube once more into segments, which are henceforth more numerous
than four. The cortical strands likewise send off contributions to the
third leaf ; they have, however, no connection with the strands of the
roots originating from the second node. The latter, as is the rule in
both Cryptogams and Phanerogams, unlike the leaf-traces, do not sub-
tend any gaps in the original central cylinder.

MONOCOTYLEDONS.

The orders hitherto discussed belong to the dicotyledonous Division of
the Angiosperms. Astely so-called is of comparatively rare occurrence
in the present group for reasons which will be subsequently suggested. In
this division, as well as in the Dicotyledons, a study of the development
of the stelar system seems to throw considerable light on its morphology.
The writer does not propose in the present memoir to more than touch
on the development of the central cylinder in the Monocotyledons, be-
cause that subject appears to him to be of great importance from the
standpoint of the phylogeny of the group at present so much in dispute,
and for that reason it seems advisable to devote a special memoir to
the development of representatives of the various moiiocotyledonous
orders.

Van Tieghem mentions Pistia stratiotes among the Aroids as possess-

i-^j,Ti^^S3K*f=&"''"'"*

28

JEFFRKY : MORPHOLOOV OF THE CENTRAL CYLINDER IN THE ANOIOSPERMS

ing an astelic central cylinder. The writer has been able to confirm the
existence of individual phlcjeotermal sheaths about the fibro-vascular
strands in this species ; but, unfortunately, in the young plants available,
the older region of the stem had already disappeared. Van Tieghem*'
does not give any account of the development of the cauline axis of
Aroids in his well-known memoir on the anatomy of the order

In Calla palustris, the young epicotyledonary stele is pithless until
after the exit of the second leaf-trace (including that of the cotyledon);
just below the point where the fibro-vascular supply of the third leaf is
given off, the central cylinder becomes tubular. The traces of the coty-
ledon and the first leaf are single ; but three bundles pass out into the
petiole of the third leaf. The three traces of the third leaf subtend
three foliar gaps in the central cylinder, which usually close at the fourth
node. A continuous fibro-vascular tube thus appears at the node of the
seedling similar to that figured by Van Tieghem^" as occurring at the
nodes of the mature rhizome. Above this node the fibro-vascular cylinder
is perforated again by gaps corresponding to traces of the fourth leaf,
which are likewise three in number. The writer has had the opportunity
of following the development of this species in specimens of not more
than six nodes. The traces of the sixth leaf are three in number, as are
those of the third, fourth and fifth. From the level of exit of the traces
of the third leaf, the fibro-vascular cylinder is astelic in Van Tieghem's
sense. The fundamental tissue forming the core of this cylinder is at
this stage not traversed by any so-called medullary strands, so that the
young central cylinder in this species closely resembles that of a Dicoty-
ledon such as Ranunculus acris. The individual strands are likewise in
the young plant collateral, and the so-called amphivasal concentric
strands make their appearance subsequently, as the writer has been
able to learn from somewhat older seedlings, which had already lost
the primary region of their stems.

The base of the young shoot in Symplocarpus fcetidus is tuberous, as
is often the case in the Aroids. Photograph 1 1, plate 8, represents part
of the ring of fibro-vascular bundles as it appears about half a centi-
metre above the base of the tuber. The individual bundles are collateral
at this stage, and by treating sections with strong sulphuric acid, the pres-
ence of a cutini/.ed phloeoterma may be demonstrated. The phlceoterma
appears in the form of sheaths surrounding the separate bundles and
hence the fibro-vascular cylinder in this phase of development is astelic.
If a section be made through the young axis in the region where the

47 Ann. Sci. Nat. Dot., 5 ser., 6 torn. Recherches sur la Structure des Aroidf^ .,.

48 Op. Cit., plate 4, fig. 5. •

'• r"

THE ANOIOSPERM8

I able to confirm the
t the fibro-vascular
jng plants available,
;d. Van Tieghem*'
f the cauline axis of
r the order,

ele is pithless until
It of the cotyledon) ;

of the third leaf is
e traces of the coty-
3 pass out into the
le third leaf subtend
ly close at the fourth
rs at the node of the
as occurring at the
bro- vascular cylinder

of the fourth leaf,
had the opportunity
:imens of not more
•ee in number, as are
of exit of the traces
: in Van Tieghem's
Df this cylinder is at
' strands, so that the
)les that of a Dicoty-
ands are likewise in
iphivasal concentric
the writer has been
:h had already lost

Hidus is tuberous, as
ite 8, represents part
! about half a centi-
)undles are collateral
phuric acid, the pres-
d. The phloeoterma
iparate bundles and
:velopment is astelic.
le region where the

JEFFREY : MOKPHOLOGV OF THE CENTRAL CYMNDRR IN THR ANGIOSPKRMS 29

tuber passes over into the narrower cylindrical stem characteristic of the
older plant, a single circular series of collateral bundles can still be
observed. If the sections through this region be treated with sulphuric
acid or with phloroglucin and hydrochloric acid, it may be demonstrated
that the bundles are no longer surrounded by individual phkeotermal
sheaths, but that a single lignified external phloeoterma surrounds ihe
whole complex of bundles. There is present in this species a state of
affairs comparable to that already described as occurring in the young
stelar system of Ranunculus rhomboideus, viz., a primitive so called
astelic condition followed by medullated monostely resulting from the
degeneration of the internal phloeoterma. The gaps between the
bundles in this stage may be compared with those subtending the
subsidiary smaller leaf-traces of many Ranunculaces ; for in the ca.se uf
the latter, the fundamental tissues of the pith and of the cortex do not
communicate through the foliar gaps becau.se these are so exiguous
that they are filled up by the pericycle flanking the adjacent fibro-
vascular segments. In Symplocarpus f<£tidus, the gaps are consequently
patent in the thick tuberous base of the primary axis, but in its narrower
subsequent cylindrical portion they are occluded by the encroachment
of the pericycle on the reduced interval between the fibro-vascular
strands. The foliar gaps are, nevertheless, morphologically indicated by
the lacunae in the fibro-vascular cylinder, filled with pericycle, which
subtend each and all of the leaf-traces, no matter how small. Not far
above the region of transition already described, the central cylinder be-
comes modified by the appearance of medullary strands and the fibro-
vascular bundles become forthwith amphivasai. Photograph 12, plate 8,
reproduces the appearance of the central cylinder in a section passing
through the cylindrical axis about half a millimetre above its junction
with the basal tuber; e is the external phloeoterma; a are the peripheral
bundles, and b are leaf-traces which have begun to run in the medulla in
the characteristic monocotyledonous fashion.

In Zea mats the epicotyledonary stele is a continuous cylinder of
fibro-vascular tissue throughout the first internode, which is generally
two or three centimetres in length. There is a single gap in the side of
the cylinder throughout the internode corresponding to the trace of the
cotyledon or scutellum. Above the second node, the fibro-vascular tube
is perforated by a large number of foliar gaps, and the leaf-traces cor-
responding to these immediately pass into the medulla, and pursue
henceforth the course so characteristic of the foliar strands in
the monocotyledons. There is a well-marked external lignified
phloeoterma and a less well-marked and somewhat interrupted internal

I

i

k'r-^

30 JEFFREY ! MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANGIOSPERMS

phlcEotcrmal sheath. The external and internal phloeotermas ap-
parently do not unite through the foliar gap corresponding to the
r-jtyledonary trace, probably for the reason that the latter is plugged
with pcricycle.

The writer believes it inadvisable at the present time to enter further
into the subject of the development of the central cylinder of the mono-
cotyledons, since it is somewhat complex and specialized, and, moreover,
presents a number of interesting features which merit a separate
consideration.

CONCLUSIONS.

As a preliminary to the description of his own observations on the
anatomy and development of the vegetative axis in the Angiosperms,
the writer has suggested certain questions to be answered by a renewed
investigation of the facts. It is now in order to discuss how far the
study of development and the re-examination of the anatomy of the
central cylinder of the Angiosperms recorded in the preceding pages,
elucidate the problems suggested in the introduction.

I. — The first question concerns the accuracy of Van Tieghem's descrip-
tion of the mode of origin of his polystelic type of central cylinder. As
has been already stated, he regards the polystelic arrangement of vas-
cular strands as the result of the repeated bifurcation of the primary
stele. If the writer's observations on the development of the central
cylinder in the Filicales as exemplified by his description of the stelar
development of Pteris aquilina, are correct, the vascular axis in Van
Tieghem's polystelic type develops as a concentric fibro-vascular tube
perforated by gaps corresponding to the points of exit of the leaf-traces.
There is consequently no indication whatever of the repeated forking
of the primary stele. In the case of the polystelic Angiosperms, the
writer's observations on the Primulaceae, Halorhagidaceae and Saxifra-
gaceae, as well as those of Gwynne-Vaughan on the Primulaceae, go to
show that here also there is no indication of the repeated bifurcation of
the epicotyledonary stele, but that the latter gives rise to a tubular
central cylinder characterized by lacunae subtending the points of t%it
of the foliar strands. It would appear that Van Tieghem has failed to
notice the extremely important influence of the leaf-traces on the con-
formation of the central cylinder in the Filicales and Angiosperms.
Had he devoted more attention to the subject of development, this
important feature could hardly have escaped his observation.

THE ANGIOSPERMS

JEFFRKY ! MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANOIOMPERMS

3'

al phloeotermas ap-
arresponding to the
he latter is plugged

time to enter further
ylinder of the mono-
ilized, and, moreover,
:h merit a separate

observations on the
in the Angiosperms,
swered by a renewed
discuss how far the
the anatomy of the
the preceding pages,
n.

n Tieghem's descrip-
central cylinder. As
arrangement of vas-
ition of the primary
iment of the central
:ription of the stelar
ascular axis in Van
c fibro-vascular tube
:xit of the leaf-traces,
the repeated forking
ic Angiosperms, the
idaceae and Saxifra-
e Primulaceae, go to
peated bifurcation of
■ves rise to a tubular
ig the points of t%it
"ieghem has failed to
af-traces on the con-
:s and Angiosperms.
af development, this
>servation.

The writer has already referred to the inadequacy of Van Tieghem's
conception of the so-called polystelic type. It seems inappropriate to
designate the young tubular central cylinder of Pteris aquilina and
Primula auricula as gamostelic, since the use of this term implies Van
Tieghem's conception of the repeated bifurcation of the primary stele
and the subsequent union of the fractions to form a stelar tube. The
study of development shows the tubular condition to be in reality primi-
tive and the writer*" has, previously, in view of that fact, suggested that
this form of central cylinder be designated siphonostelic.

The studies of Gwynne-Vaughan and of the writer on the Primulacca?,
Halorhagidaceae, and Saxifragaces seem to show that the siphonostelic
type of central cylinder, with internal phlotim, is not to be regarded as
primitive in the case of the Angiosperms, but that it is derived from a
siphonostelic type in which there is primitively no internal phloem. This
conclusion is justified by the fact that the young stelar tube is some-
times actually without internal phloem, e.g.. Primula japonica, P.
farinosa, and Gunnera scabra, and by the fact that, even when the young
stelar axis is concentric from the first, the leaf-traces retain the
apparently ancestral collateral type, e.g., Primula Auricula. Gwynne-
Vaughan attributes the appearance of internal phloem in the central
cylinder of certain Dicotyledons to an effort to make up for the loss
of a cambium, a feature correlated with so-called polystely.

II. — The writer's study of the development of the so-called astelic
central cylinder of certain Ranunculaceae and Nymphseaceae shows that
in this type, the course of events is practically the same as that in the
so-called polystelic axis, vis., that the epicotyledonary stele becomes in
the young plant a tube interrupted by foliar gaps. The tubular
character of the stelar system may be subsequently disguised by the
increasing complexity of the arrangement of the fibro-vascular strands,
but in all such cases the collateral stelar tube may be recognized in the
young axis. The writer has not been able to find any evidence in the
facts of stelar development in seedlings, supporting Van Tieghem's
statement that his astelic type originates by the separation of the
phloeoterma into segments, which unite around the individual bundles
Neither do his observations accord with the more recent description
given by Van Tieghem'" of the astelic type, in which he states that
astely or schizostely as he now prefers to call it (adopting a suggestion
of Strasburger's), originates by the stele breaking up into as many
meristeles as it contains bundles. The young stele in these cases is

49. Trans. Brit. Ass. Adv. Sci., 1897, p. 869.

50. ^Mmentsde Botanique, p. 179. . -

I

"T

3a JEPPRBV : MORPHOUHJY OV THE CKNTRAL CVI.lNnEK IN THK ANC.IOSPERMS

always a continuous hollow cylinder of fibro-vascular tissue, bounded
internally and externally by phloeotermal sheaths, which communicate
through the foliar lacuncX-, and is in fact primitively of the conformation
described by Van Ticghcm as gamodesmic or gamomeristclic. The use
of these terms, however, seems inappropriate, since they imply the union
of bundles or meristeles originally separate. The writer has ruggested
the term siphonostelic as fitly describing the stelar type occurring in the
stem of P/ens aquilind or Primula farmosa. The question may now be
asked whether this term should not also beapplied to thevery similar stelar
conformation found in the Ranunculacc;e and Nymph.-caceie. The most
striking difference between the stelar systems of the two types is the
absence of an internal phloem in the case of the Nymph;eace;e and
Ranunculace.-t. But Gwynne-Vaughan has called attention to the fact
that the collateral fibro-vascular strands of the Nymphcvace.t often
become concentric, an observation which the present writer has been
able to confirm. On the other hand, the siphonostelic central cylinder
ot" Primula japonica and P.farinosa starts as a collateral stelar tube,
only subsequently becoming more or less completely conv <..ntric. In the
floral axes of these species, moreover, both the internal phloem and the
internal phloeoterma disappear, giving rise to a central cylinder, which,
if its mode of origin had not been followed, would be regaroed as
medullated monostelic. The flowering stems of all the described species
of so-called polystelic Primulas have the same peculiarity, It has been
shown in the earlier part of this essay that the stelar strands oiParnassia
palustris, within an interval of a fraction of a millimetre, may be
successively collateral and concentric, whi ■; in P.parviflora the cauline
strands are collateral and the leaf-traces concentric. The two types of fibro-
vascular strand pass imperceptibly into each other in the Angiosperms,
and, as will be subsequently shown, also in the Gymnosperms and the
Vascular Cryptogams. It would appear from the study of development
that Van Tieghem's polystelic and astelic types are essentially the same,
inasmuch as the stelar system in both cases is primitively a tube with
foliar lacunai. Further, the presence or absence of internal phloem
appears to be a matter of slight morphological importance.

HI. — We may now pass to the question of the morphology of the
medullated fibro-vascular axis, bounded by an external phloeoterma
only. It has already been pointed out that Van Tieghem regards the
central cylinder in this type as derived from the pithless stele of the
hypocotyl, by dilatation and the formation of a parenchymatous medulla,
differing morpholog"cally from the fundamental parenchyma outside the
stele. It has also been mentioned that Van Tieghem considers that his
medullated monostelic type of central cylinder, and his astelic type, may

•>i<.iil a<<tr'''"i'^- '-•i^-<''

HB ANCilOSPKRMS

JEFFREY : MORPllOLOfiV OF THR CRNTH.Vr I'Vl.l.VrtRR IN TIFF AS

ar tissue, bounded
.rhich communicate
if the conformation
neristclic. The use
ley imply the union
riter has ruggested
■pc occurring; in the
lestion may now be
levery similar stelar
i.taceie. The most
le two types is the

Nymphieaccii; and
ittention to the fact
Nfymphteace.t often
nt writer has been
elic central cylinder
illateral stelar tube,

coni f.ntric. In the
nal t>Klocm and the
;ral cylinder, which,
aid be regaroed as
he described species
iarity, It has been
strands of Pamassia
millimetre, may be
xrviflora the cauline
he two types of fibro-
in the Angiosperms,
mnosperms and the
udy of development
essentially the same,
litiveiy a tube with

of internal phloem
irtance.

: morphology of the
ixternal phloeoterma
rieghem regards the
pithless stele of the
tichyraatous medulla,
enchyma outside the
sm considers that his
his astelic type, may

co-exist in tliffcrent species of the same genus. For exatnpli Kaniim
aquatilis is, according to his view, astelic, and A', abortiius is, on the
other hanil, mcdullatcd nionustclic. htrasbiugcr has (luestioned the
possibility that the pith in different species of Ranunculus should be
sometimes cxtrastelar (in the astelic type), and intrastelar (in the
monostelic type). He prefers to consider that the ineilulla is always
intrastelar, and that, consequently, the individual endcniermal sheaths
in the species of Ranunculus, wlurc they occur, arc different niorpho-
1 igically from the [)hlceoterina, which he supposes surrounds the
complex of bundles in the so-called mcdullated monostelic types.

The writer's examination of the development of a number of species
of Ranunculus and Anemone has shown that in the young axis the
stelar system possesses an internal phlceoterma which is continuous
with the external phloeoterma through the foliar gaps, and is therefore
of the same morphological value. In some species the internal phkjeo-
terma disappears in the older region of the stem, e.g., R. abortivus and
A', rhomboideus, and the stelar axis then appears to be monostelic. A
.study of its development, however, shows that it is primitively astelic.

This view of the matter quite removes the difficulty of having to
regard the pith in some species of Ranunculus and Anemone as intra-
stelar and in other species as extrastelar, because their ontogeny
shows that the pith is in both ca.ses extrastelar, i.e., ordinary
fundamental tissue which has been enclosed by the stelar tube. In the
older stem of various species of Ranunculus and Anemone, both
external and internal phlceotermas disappear when there is much
sclerenchymatous tissue present in the vascular .system. It is probably
not going too far to state as a result of the study of development that
in the genus Ranunculus the pith is throughout extrastelar. In
consequence of his detailed anatomical study of the order Ranunculaceae,
Marid" has come to the conclusion that the genus Ranunculus is the
stock from which all the other genera of Ranunculacea; have taken
their origin. It will probably not be an ea.sy matter to demonstrate
the existence of an internal phloeoterma even in the young stems of
the various ranunculaceous genera, especially where there is much
sclerification of the central cylinder or much secondary growth, since
even the external phloeoterma under these circumstances is nearly
always with difficulty distinguishable. It is not unlikely, however, that
the internal phloeoterma will be found to persist in the more conserva-
tive vascular system of the floral axes. In any case, there appear to be

51 Op. Cit., p. io8.

LofC.

wu^imammm^-

ta^fe-i^-atiii.

1>

JUrrHRV I MO«PHOI.OOV op tup. CRNTRAI, rVtlNPFR IN TUV. .\Nr.li>SI'i:RMH

better reasons for rej^iinliiit; the Riiminculuceii.' throughout as astclic, in
Van Tiq,'hc'm's sense, than as entirely ineduUatccI nioiiostelic, according
to Strasbur^,'er, or in part inedullatcd nionostelic, and in part astclic,
accordin^j to Van Tic|.,'hcni.

If the central cylinder of the Ranunculacciu be throughout collateral
siphonostclic in the sense described by the writer, it would seem
necessary to extend that conception of the morphology of the central
cylinder to the rest of the dicotyledonous An^dosperms. It appears
hit,'hly probable that an exhaustive study of sccdlinj^s and lloral axes of
rc|)resentatives of the various orders will result in the disci>very of a
number of facts favorable to this view, anrl an invcstifjati. n in this
direction is in pro^jrcss.

The writer's observations on Monocotyledons point to the same
[general conclusion, althoujjh here the reduction in size of the foliar
iacun.i- and the hij^h development of the pericyclic mechanical tissues
which has brou^dit this division into special prominence in connection
with Schwendener's mechanical theories, make the demonstration of an
internal phlreotcrma much more difficult. The Aroi<ls have been
chosen to illustrate preliminarily the fundamental astely of the
Monocotyledons. Qi/Za palitstris is throughout astelic in Van
Tieghem's sense ; Symplocarpus fcetidus is obviously astelic in the
young stem, while in the older axis it is apparently medullated
monostelic, on account of the degeneration of the internal phlceoterma.
We may preliminarily assume that the Aroids have throughout a
pith derived from the fundamental tissues. This conclusion may be
extended to the other Monocotyledons and it may be further stated
that the primitive stelar condition in the Monocotyledons is tubular,
and that the central cylinder is interrupted by foliar lacunic correspond-
ing to the leaf-traces. The peculiar cauline course of the bundles in the
Monocotyledons is not primitive, since it does not appear in the young
stem.

Taking into consideration all the facts derived from a study of the
Angiosperms, the general statement may be made that the primitive
type of stelar system in the group is a hollow fibro-vascular tube with
gaps corresponding to the leaf-traces. This type of stele may appropri-
ately be called siphonostclic. In siphonostelic axes the stelar tube may
have both internal and external phloem, and this case may conveniently
be described as amphiphloic. Wl er^ the internal phloerp is absent, as
is the case in the majority of the Angiosperms, the stelar tube is ecto-
phloic. The pith is always to be regarded as merely an included portion
of the fundamental tissues.

y ^^.i'^ -r-m^^wiafra

itwfjiwi'wii

UK \N(1ii>spi:rmh

JRPMrv ! MORPHOI.OOV Of TIIK C'KNTR \l. CVI.INnKR IN TIIR ANfllORPRRMH .U

^;h()Ut as astc'lic, in
jiiostclic, accordinn
,iul in part ustclic,

rouphout collateral
cr, it would seem
lof;y of the central
perms. It appears
s and floral axes of
the disci>very of a
vesti^atii 11 in this

joint to the same
1 size of the foliar

mechanical tissues
nence in connection
Icmonstration of an

Aroi'ls have been
ital astely of the
it astelic in Van
usly astelic in the
arently medullated
iternal phlceotcrma.
have throughout a
conclusion may be
ly be further stated
tyledons is tubular,
lacuniE correspond-
)f the bundles in the
ippear in the young

from a study of the

that the primitive

3-vascular tube with

stele may appropri-

the stelar tube may

;e may conveniently

ahloerp is absent, as

e stelar tube is ecto-

• an included portion

In the reduced central cylinder of certain an^'iospermous species of
aciiiatir or amphibious habit, the stelar system has become contracted
on accoijr-t of the correlated dc^jeneracy of tb.e vascular elements. In
these c.iso, the parenchymatous core in the centre of the stele, tx, that
of P(wnassia paliistris, represented in phot()nrai)h 25, plate 11, is to be
regardc'l as included pcricyclc, morphologically similar, lor example, to
that fou, ' in the >'<>////(,'• fibro-vascular axis of I'riiiiiilii fariuout. The
contracti I central cylinder in such cases, is none the less to be regarded
as essentially siphonostelic, although it has no true medulla.

I v.— We may now ask if the study of the development of the central
cylinder in the Angiosperms supi)lii , any morphological facts which arc
of |)hylogcnctic value. Attention has already been directed to the
primitively tubular character of the central cylinder throughout the
Angiosperms. It has been furthermore noted that the stelar tube is
characterized by foliar gaps corresponding to the traces of the leaves.
It is of interest to discover whether the occurrence of foliar lacuna; is
constant, or whether it is in any way innuenccd by environment. The
Cactaceie and the amentiferous genus Casuarina present perhaps the
most extreme cases of foliar reduction among the Angiosperms, and it
is of interest to discover whether the foliar gaps are obsolete in
these examples. The writer has satisfied himself by examination of
species of Opuntia and Ccreus, of the accuracy of Ganong's'' statement
as to the occurrence of gaps in the fibro-vascular system of the Cactace.e,
corresponding to the extremely reduced leaves. The .stelar system of
seedlings of Casuarina equisetifolia grown from seeds obtained from
Kcw, is al.so characterized by the presence of well-marked foliar gaps.
It may consequently be assumed that the occurrence of foliar gaps is a
constant characteristic of the Angiosperms.

The po.ssession of an essentially tubular stelar system interrupted
invariably by foliar gaps appears to be a palingcnetic feature of the
Angiosperms. In this group there are so few anatomical characters which
can rank as phylogenetic criteria that the demonstration of an additional
one may be expected to contribute something to the solution of the
extremely difficult problem of its origin.

In an earlier memoir,'^' the writer has called attention to the value of
a study of the stelar system in connection with the phylogeny of vas-
cular Plants. It would indeed be strange if an apparatus so character-
istically separating them as a. whole from the lower Cryptogams did not

5a Beitrag. Z. Kennl, d. Morph, u. Biol. d. Cacteen. Inaugural. Dissertation, p 13 etc.
53 Mem. Boston Soc. Nat. Hist., Vol. 5, No. 5.

T tmv\"m rn.Wiir -if^W:'^

36 JEFFREY : MORniOLOr.V OF THE CENTHAI. CVLINOFR IN THE ANGIOSPERMS

manifest some features of phylogenetic importance. The writer has
pointed out that siphonostcly exists under two modifications among the
vascular Cryptograms, viz., siphonostcly, in which the gaps of the stele
correspond to leaf-traces ; and siphonostcly, in which there are no foliar
gaps, but in which the stelar lacunai correspond to branches. He has
designated the former type phyllosiphonic ; the latter, cladosiphonic.
The use of this distinction, together with all other available character-
istics, results in the placing of the Lycopodiales and Equisctales near
one another in the system as groups which are invariably cladosiphonic.
The Filicales, Gymnosperma:, and Angiosperma;, on the other hand,
arc uniformly phyllosiphonic. The writer does not intend to go further
irvto this matter in the present memoir, but it may nevertheless be sug-
gested that there are two distinct primitive groups of vascular plants,
viz., the Lycopsida and the Pteropsida. To the former group belong
the Lycopodiales and Equisctales ; to the latter, the Filicales, the
Gymnospermai, and Angiosperma;. The validity of the classification
indicated above, in the case of Lycopsida, has already been discussed by
the writer,''^ in his memoir on the genus Equisetum. The considera-
tions which favor the setting up of the other great alliance can be
more advantageously examined subsequently.

V. If the writer has correctly interpreted the anatomical facts
described and figured in the present memoir, the morphological ideas
of Van Tieghcm can no longer be accepted in full. It is the great
merit of that distinguished botanist to have recognized in so large a
measure the essential unity of the fibro-vascular system of plant axes.
That he should have gone too far in the direction of unity in the case of
the so-called medullated monostelic type, and not far enough in
his polystelic type, is to be explained by the comparatively slight
attention given by him to the subject of development. Strasburger,
in respect to monostely, has gone even further than Van Tieghem,
for he unites Van Tieghem's astcHc type with the medullated
monostelic, and regards the medulla in both cases as intrastclar
parenchyma. An examination of the anatomy of the young axis
apparently brings us back to the standpoint of De Bary in regard
to the morphology of fibro-vascular .strands. The older anatomist,
however, on account of his intentional neglect of ontogeny, appears
to have completely overlooked the morphological unity of the cauline
fibro-vascular apparatus of the Angiosperms.

Since it is impossible to investigate experimentally any but the

s« Mem. Boston. Soc. N.it. Hist., Vol. 5, No. 5.

a "^ ■' —rt— "■-."-'*•'' ""i ■ "**!'"

rHE ANGIOSPERMS

e. The writer has
fications among the
ic gaps of the stele
I there are no foliar

branches. He has
itter, cladosiphonic.
available character-
d Equisetales near
iably cladosiphonic.
on the other hand,
intend to go further
nevertheless be sug-

of vascular plants,
irmer group belong
, the Filicales, the
of the classification
y been discussed by
im. The considera-
eat alliance can be

i anatomical facts
morphological ideas
ill. It is the great
Tnized in so large a
stem of plant axes.
" unity in the case of
not far enough in
omparatively slight
ment. Strasburger,
han Van Tieghem,
th the medullated
:ases as intrastelar
of the young axis
De Bary in regard
he older anatomist,
ontogeny, appears
unity of the cauline

ntally any but the

JEFFREY : MORPHOI-OC5Y OF THE CENTRAL CYLINDER IN THE ANC.IOSPERMS

.17

smallest morphological problems, all interpretations of the greater
questions of morphology must ever remain more or less hypothetical.
The persistent recurrence of a tubular stelar system throughout the
various groups of vascular plants suggests that the tubular arrangement
of the primitive skeletal tissues is often an advantage. The nature of
the advantage is apparently not far to .seek. One important function
of plant axes is to afford support to their appendicular organs, and we
may regard the tubulization of the stele as an adaptation for this
purpose. Where the mechanical function is taken on by the extra-
stelar tissues, siphonostely is often absent or degenerate. Further,
those organs of plants which are normally supported by the soil, viz.,
the roots, are not primitively siphonostelic at all, and in the com-
paratively rare cases where their central cylinder is medullated, its
parenchyma is derived from the pith of the stem. The truth of the
latter statement will be more apparent when the development of the
stem in certain Filicales has been described. It would consequently
appear that the tubular fibro-vascular cylinder, which is so characteristic
of the cauline axis of the Angiosperms, is the result of the operation of
mechanical causes, and that the anatomical peculiarities which distinguish
the primary central cylinder of the root from the stelar tube of the
shoot are primitive features, retained undisturbed by the mechanical
influences acting on the stem.

SUMMARY OF RESULTS AND CONCLUSIONS.

1. The polystelic type of Van Tieghem is not characterized by the
repeated bifurcation of the epicotyledonary stele, but there is primitively
in the young stem of this type a tubular concentric stele with foliar
gaps subtending the points of exit of the leaf-traces.

2. The astelic type of Van Tieghem does not result from the separa-
tion of the epicotyledonary stele into its con.stituent bundles ; for in the
young so-called astelic axis, there are no bundles present at all, but a
collateral stelar tube with foliar gaps subtending the leaf-traces, through
which the internal and external phloeotermal sheaths communicate.

3. The medullated monostelic type of Van Tieghem does not originate,
as he states, by the dilatation of the epicotyledonary stele and the forma-
tion of an intrastelar pith ; for the writer's observations show that in favor-
able cases, both among the Dicotyledons and the Monocotyledons, the so-
called medullated monostelic central cylinder of the older stem may be
seen to be derived from the so-called astelic condition of the young
axis, by the degeneration of the internal phloeoterma.

t1lir.tfa^i*iMiir,

38 JKFFREV : MORniOLOGY OK TIIE CENTRAL CYLINDER IN THE ANfilOSPERMS

4. Van Tieghem's three types of central cylinder indicated above are
all merely modifications of a single type, which has been designated
by the writer, siphonostelic. In this type the central cylinder is
primitively a fibro-vascular tube with foliar lacun;v opposite the points
of exit of the leaf-traces. In the so-called polystclic modification, the
central cylinder has internal, as well as external phloem, and may be
described consequently as amphiphloic. In the astelic type of axis so-
called, the internal phlocim is absent and the central cylinder is accord-
ingly to be designated ectophloic. The medullated monostelic type of
Van Tieghem is derived from the last-named by the degeneration of
the internal phkcoterma or endodermis.

5. The siphonostelic type of central cylinder as defined above is
probably to be regarded as the result of the mechanical strengthening
of the cauline axis to enable it to support the palingenetically
large leaves which are characteristic of the Angiospermai, Gymno-
sperma:, and Filicales. In these three groups, the siphonostelic, fibro-
vascular cylinder is invariably distinguished by the presence of gaps
corresponding to the points of exit of the leaf-traces, and, in this fea-
ture, offers a marked contrast to the tubular central cylinder of the
Lycopodiales and Equisetales, in which there are no foliar lacuna;, but,
on the contrary, gaps subtending the branches.

6. A study of the development and structure of the fibro-vascular
apparatus of the various groups of vascular plants is likely to throw
considerable light on their phylogeny and to elucidate the causes of the
morphological differences in the structure of the central cylinder of root
and shoot.

For many kindnesses in the matter of supplying material for this
research, the writer wishes to express his obligations to the Director of
the Royal Gardens, Kew ; to the Director and Assistant Director of
the Botanical Gardens of Harvard University; to Dr. D. H. Scott,
Honorary Director of the Jodrell Laboratory, Kew, and to Mr. Jackson
Dawson, of the Arnold Arboretum of Harvard University. He is
specially indebted in this respect to Prof. G. L. Goodale, Director of the
Botanical Gardens of Harvard University. Lastly, the writer owes not
a little to his assistant, Mr. R. B. Thomson, B. A., for help in securing the
seedlings necessary for the present investigation.

IE ANfilOSPERMS

JEFFREY : MORPHOLOGY OF THE CENTRAL CYLINDER IN THE ANGIOSPERMS

.19

ndicated above are
3 been designated
entral cylinder is
ipposite the points
; modification, the
loom, and may be
ic type of axis so-
cylinder is accord-
monostelic type of
\e degeneration of

1 defined above is
lical strengthening
le palingenetically
osperma:, Gymno-
siphonostelic, fibro-
; presence of gaps
;s, and, in this fea-
•al cylinder of the
5 foliar lacunae, but.

the fibro-vascular

is likely to throw

te the causes of the

tral cylinder of root

g material for this
s to the Director of
ssistant Director of
Dr. D. H. Scott,
and to Mr. Jackson
University. He is
dale, Director of the
the writer owes not
help in securing the

DESCRIPTION OF THE PLATES.

PLATE VII.

Photograph i. — Transverse section of a younjif stem of /Vt-w ai/«///'mi, at the point of
origin of the second leaf. (X 40.)

Photograph 2. — Transverse section of the older stem of /'. a(/uilitta ; /' and /' are
leaves ; >• is a root. (X 25.)

Photograph 3. — Transverse section of the stem of P. ai/iiilina above the reunion of
forking of the stelar system j a and /> are the two cylindrical
steles; / is a leaf-trace. (X 15.)

Photograph 4. — Transverse section of the young horizontal rhizome of /'. aquilina.
(X 25.)

Photograph 5. — Transverse section of the older horizontal rhizome of P. aquilina.
(X 25.)

Photograph 6. — Transverse section of a still older region of the same. (X 25.)

PLATE VIII.

-Transverse section of young stem of Primula farinosa. (X 40.)

-Transverse section of the older stem of P. farinosa ; P .-ind /' indicate
the foliar gaps ; ;-• and r' are roots. (X 25.)

-Transverse section of the same stem with same lettering as pho-
tograph 8. (X 25).

— Transverse section of the young stem of Nelumhium luteum, at the
point of origin of the cotyledons. (X 25.)

-Transv.irse section of part of the tuberous region of the stem of
Symblocarpus fatidus, showing the so-called " astelic " condition.
(X 25.)

Photograph 7.-
Photograph 8.-

Photograph 9.-

Photograph 10.

Photograph ii

Photograph 12. — Transverse section of cylindrical portion of the stem of the same,
showing the "medullated monostelic " condition; e, phlcEO-
terma; a, peripheral strands; b, foliar traces. (X 20.)

PLATE IX.

Photograph 13. — Transverse section of the young stem of A'anunculus acris. (X 25.)

Photograph 14. — Transverse section of the oUlcr stem of A'. oiz-jV. (X 25.)

Photograph 15. — Transverse section of the older young stem of A', abortivus. The
internal phloeoterma is still perceptible. (X 25.)

Photograph 16. — Transverse section of the older young stem of A", rhomboidcus. The
internal phloeoterma cannot be distinguished. (X. 25.)

m

I'M-

40 JEFFREY ! MORPHOLOGY OF THE CENTRAL CYLfVDER IN THE ANGIOSPERMS

Photograph 17. — Transverse section of the young stem of /'. rhomboUeus.
internal plilcsoterma is still visible. (X. 25).

The

Photograph 18. — Transverse section of the young stem of K, rhomhoukus, showing
the communication of internal and external pliloeotermal sheaths
through the foliar lacuna. (X. 25.)

Photograph 19.
Photograph 20.

PLATE X.

—Transverse section of the young stem of Anemone pennsylvanica,
showing the presence of an internal phlceoterma. (X. 25.)

— Transverse section of the older young stem of A, pennsylvanica,
showing the overlapping of the foliar gap. (X 25.)

Photograph 21. — Transverse section of the young stem of Brasenia purpurea, showing
the primitively tubular condition of the stelar system. (X. 20.)

Photograph 22.
Photograph 23.
Photograph 24

-Transverse section 01 the young stem of Nuphar luteum, var.
Kalmiannm, showing the tubular stele. (X 25.)

— Transverse section of the older young stem of Nuphar advena,
(X 25.) ■::■

-Transverse section of the young epicotyledonary stctar system of
Nelumbiuni luteum ; a, primitive stelar strands ; /, cortical
bundles. (X 25.)

. - PLATE XL ,.

Photograph 25. —Central cylinder of Parnassia palustn's. (X 45. )
Photograph 26,

Photograph 27.-
Photograph 28.-
Photograph 29.-
Photograph 30.-

-" Polystelic " phase of central cylinder of P. palustris, /' and /" are
leaf traces. (X 12.)

-Inferior region of " astelic " portion of central cylinder of P.
palustris, (X 45.)

-Transverse section of the stele of P. palustris, at point of origin of a
leaf-trace (X 12.)

-Transverse section of " astelic " central cylinder of P. parviflora,
g^ and g' are foliar gaps ; / is a leaf-trace. (X 25.)

-Transverse section of central cylinder of the young stem of Hepatica
triloba. (X 40.)

HE ANGIOSrERMS

/'. rhomboidcus. The

25)-

', rhomhoideus, showing
al phlccotermal sheaths

Anemone pennsylvanica,
ioterma. (X. 25.)

ini of A, pennsylvanica,
p. (X 25.)

senia purpurea, showing
;elar system. (X. 20.)

f Nuphar luteuin, var.
(X 25.)

em of Nuphar advena.

onary stclar system of
' strands ; /, cortical

5-)

palustris, /' and /" are

central cylinder of P.

's, at point of origin of a

irlinder of P. parviflora,
,ce. (X 25.)

I young stem of Hepatica

..dftli-i

|H.\.\SA< lluNS UK nil. (ANAKIAN InhIIII IK. \i)|. \'l

I'l.AII. MI

/ -'■

Hq.l

Hif.l

n,,:>.

Fu,.l.

Ik'utWt f^WJV •».^*!tf.

-'*miMmmml»mimi^tUt.\ m

r

I'l All. VIII.

ir^r' '•" ^*»* t, flf U-.

liiANsAi rinNsiii nil ( Waiiia.s Inxhii i |. . \(ii \|

I'l \ii. l\

h),.n

hiiJi

/•/(/./»

liilJli

h,f.n.

ih,JK

»,«it./ *^s«-/ ^..'••tf

C55*»*««sW»»W*^*^SBl.5»»aflB

B^ewwiitwggTCT'W'P

|H.\.NH.V llllNH (l> lllh t VNAIIIAN I.NXIIII It.,, Vul. \l

I'l.AII. \

^^tv

<c: ;.<,w

>iiA:'i'.

-• "

fiifU

JbU^tvpr "rpinv *. M**t)^

I'l \ii M

>♦'

■■•7ia»S«It™*««.'»;*'

^rri^igtfta'ia^'f f«iBiegaawgwMB»w «»n<>g*iwiwwf»giy iMh-f^^w'-,^