Early  Journal  Content  on  JSTOR,  Free  to  Anyone  in  the  World  
This  article  is  one  of  nearly  500,000  scholarly  works  digitized  and  made  freely  available  to  everyone  in  
the  world  by  JSTOR.    

Known  as  the  Early  Journal  Content,  this  set  of  works  include  research  articles,  news,  letters,  and  other  
writings  published  in  more  than  200  of  the  oldest  leading  academic  journals.  The  works  date  from  the  
mid-­‐seventeenth  to  the  early  twentieth  centuries.    

 We  encourage  people  to  read  and  share  the  Early  Journal  Content  openly  and  to  tell  others  that  this  
resource  exists.    People  may  post  this  content  online  or  redistribute  in  any  way  for  non-­‐commercial  
purposes.  

Read  more  about  Early  Journal  Content  at  http://about.jstor.org/participate-­‐jstor/individuals/early-­‐

journal-­‐content.    

JSTOR  is  a  digital  library  of  academic  journals,  books,  and  primary  source  objects.  JSTOR  helps  people  
discover,  use,  and  build  upon  a  wide  range  of  content  through  a  powerful  research  and  teaching  
platform,  and  preserves  this  content  for  future  generations.  JSTOR  is  part  of  ITHAKA,  a  not-­‐for-­‐profit  
organization  that  also  includes  Ithaka  S+R  and  Portico.  For  more  information  about  JSTOR,  please  
contact  support@jstor.org.  
VOLUME LVII NUMBER 3

THE

BOTAN ICAL GAZETT

MIARCH 1914

THE ANATOMY OF OPHIOGLOSSUM PENDULUM

CONTRIBUTIONS FROM THE HULL BOTANICAL LABORATORY i83
LOREN C. PETRY

(WITH SIXTEEN FIGURES)

The section OPHIODERMA of the genus Ophioglossum,as first

separated by PRANTL (6), is distinguishedfromEUOPHIOGLOSSUM
by the numerousvascular strands in the base of the petiole, and
fromCHEIROGLOSSA, which shares this feature,by the occurrence
of a single spike on each fertileleaf. As originallydescribed,the
section was composed of the single species 0. pendulumL.; 0.
intermedium Hook. was consideredto be a youngformof 0. pendu-
lum, and was included with that species. In I904, BOWER (2)
described the new species 0. simplex Ridley, fromSumatra, and
froman examination of its vascular system showed that it is a
member of this section. He furtherreestablishedthe species 0.
intermedium Hook. on the basis of its intermediatepositionbetween
0. pendulumand 0. simplex.
Of the three species constitutingthe section, 0. pendulumis
best known. BOWER (i) has described the development of the
fertilespike and sporangia,and in a later paper (2) has given a
briefaccount of its anatomy. This paper deals merelywith the
vascular systemof the leaf and its insertionupon that of the stem.
The investigationhere describedhas been undertakenwith a view
to supplementingthis account of the anatomy of the species.
I69
170 BOTANICAL GAZETTE [MARCH

Material
A large numberof entireplants was collected by Dr. CHARLES
J. CHAMBERLAIN in November I91 I, in NorthernAustralia. The
exact localityis near Babinda, in the Cairns districtof Queensland,
at a distanceof about 25 miles fromthe coast. Occasional clusters
of plants were found near the ground, but more usually they
occurred at a considerable height, up to 30 meters. In almost
all cases they were growingin the masses of humus accumulated
by large plants of Polypodiumrigidulum; Lycopodiumphlegmaria
was a commonassociate. Plants of this species were also observed
on StradbrokeIsland, near Brisbane,but no collectionsweremade.
The groupsof plants here were at a less distance fromthe ground;
they were commonlyattached to plants of Platyceriumalcicorne
and P. grande.
In the material secured at Babinda, the largest leaves had
attained a length of about I.5 meters. However, some of the
plants on StradbrokeIsland had leaves nearly 2 meterslong, and
Mr. WM. GIBSON, a collectorfamiliarwith that region,reported
a specimen with leaves measuring 2.7 meters in length. The
largestspikes seen measured30 cm. in lengthand I. 2 cm. in great-
est width. Many cases of branchingor lobing of the spikes were
observed; probably one-tenth of the specimens showed some
variation fromthe simple form.
This supplyofentireplantswas supplementedby a largenumber
of rhizomes,root tips, buds, and spikes collected by Dr. W. J. G.
LAND on the island of Tutuila, Samoa, in October i9I2. These
plants, togetherwith many fernsand occasional plants of Lycopo-
dium phlegmaria,occurredin the root masses of Asplenium'nidus,
at a heightof 2-IO meters; a large clusteris shownin fig.i. The
leaves reached a lengthof 2 meters; branchingof the leaf was not
uncommon,but branchingor lobing of the spike was relatively
rare. The largest spike collected measured 5i cm. in lengthand
I.3 cm. in width; sporangia occurredalong 45 cm. of its length.
Several spikesmorethan 40 cm. in lengthwere found.
The Australianmaterial was preservedin 6 per cent formalin;
that securedin Samoa was killed in 50 per cent alcohol and 6 per
cent formalin. Even the largest rhizomescut readily in paraffin
I9I4] PETRY-OPHIOGLOSSUM PENDULUM I7I

at IO-25 A, and completeserieswereeasilyobtained. The safranin-

anilin blue stain combinationgave the best results.

FIG. I.-phioglossUM pezldthlntm, uponCitrussp.; Tutuila,Samoa;

epiphytic
photographby Dr. W. J. G. LAND.
I72 BOTANICAL GAZETTE [MARCH

The root
The root stele at the point of attachmentto the stem cylinder
varies fromdiarch to pentarch; triarchand tetarcharrangements
of the xylemare commonest. In passing fromthe base of the root
toward the tip, the number of protoxylem points frequently
increasesby the splittingof one of the originalprotoxylemstrands;
in one case, the disappearance of a protoxylemstrand was noted.
In a few cases only,a hexarcharrangementwas found; these roots
were all very large. In generalthereis a definiterelationbetween
the size of the root and the numberofprotoxylemstrands.
Each phloem mass is a layer i or 2 cells in thickness,separated
from the xylem by 3-5 layers of parenchyma. It is easily dis-
tinguishedby the somewhat thickenedwalls and the conspicuous
"proteid granules" adheringto the walls. In the apical regionof
the root,the protophloemcan usually be distinguishedconsiderably
in advance of the protoxylem. When a protoxylemstrand splits,
the phloem between the branchesof the split strand connectswith
the phloemmasses on eitherside.
There is no pericycliclayer; the phloem abuts directlyupon
the endodermis(fig..Io), whichcan be readilydistinguishedby the
suberizedband upon the radial walls; it resemblesthe endodermis
of Botrychiumvirginianumin all particulars. Two or three cells
of parenchymaseparate the protoxylemfromthe endodermis.
Two general regionsof the cortex about equal in thicknessare
distinguishable. In the inner region, the cell walls are heavily
thickenedby cellulose; simple circularor oval pits occur. In the
outer region,the thickeningof the cell walls is much less and gives
the reactions of pectin compounds. Intercellularspaces of con-
siderable size occur in the innerregion; in the outer region,these
are much smaller and become filledby the substance which con-
stitutesthe thickeningof the cell walls. The endophyticfungus
which occurs commonly in the roots is confined to the outer
region.
The outer walls of the surfacecells are remarkablythickened,
as shown in fig. 2. The thickeningmaterial is not cellulose, but
gives the reactionof pectin compounds. The thickeningmay con-
tinueuntil the lumenof the cell is almost filled; even in these con-
I9I4] PETR Vi-OPHIOGLOSSUM PENDULUMI I73

ditions the protoplasm remains very active. When cells of the

epidermallayer are killed, the outer walls of the cells immediately
below begin to thicken in this way. This
may occur in any of the cells of the outer
region of the cortex upon destructionof the
overlyingcells.
Roots originatein the meristematicregion
of the stem by the formationof an apical cell FIG,. 2.-EpidCrmal
in the firstlayer of cells outside the phloem. cell of root; X 236.
Fig. 3 represents this origin as seen in a
longitudinalsection of the stem. As shown,differentiation of the
phloem has proceeded to withina few cells of the position of the
root meristem. The growthof the rootproceedsby the segmenta-
tion of the apical cell, which
is tetrahedral in form; in
slowlygrowingroots the seg-
mentationis irregular,and a
0;_Ag< O ) p definiteroot cap cannotbe
distinguished. The first
xylem elements are smaller
in cross-section than those
formedlater; they are true
J5A// < / _ |tracheids with walls reticu-
lately thickened. In a few
cases of more rapidly grow-
3.
_
> ing roots, the thickeningof
the walls approaches the
annular marking. This char-
acter of the protoxylemindi-
cates the slow growthof the
roots. The significationpro-
ceeds slowly toward the
FIG. 3.-Origin of root, as shown in centerof the stele,and xylem
sectionof a bud: ph, phloem;
longitudinal
XX26
236. is oftenfoundin
parenchyma
the centerof the stele oflarge
roots at a distance of 3-4 cm. fromthe tips. There is no secondary
thickening of the xylem mass. In a few cases a medullated
I74 BOTANICAL GAZETTE [MARCH

conditionoccurred,as shownin fig.4; the pith is not xylemparen-

chyma,but is composed of cells resemblingthose of the cortex; the
opening throughthe xylem always occurs, but the endodermisis
not interrupted. This occurs only in rootsupon whichbuds have
developed stems; the open portion of the xylemmass is upon the
upper side of the root. This condition persists for only a short
distance from the bud toward the tip of the root; the gap soon
closes and a solid xylem mass is again formed. This peculiar
rootstele formis directlyrelated to the formationof the stem stele
frombuds upon the roots,as describedlater.
The branchingof the root is strictlymonopodial. The lateral
roots grow at an angle of 6o-90o to the main root; two or three
branch roots oftendevelop at one point. These are not restricted
in theirgrowth,as in Helminthostachys,
e"> <: but may reach as great length as the
primaryroots.
Immediatelybeforea branchis given
off,the phloem of the main stele spreads
laterally between the endodermis and
X h the protoxylem,and formsa continuous
FIG. 4.-Medullated root sheathabout the xylem. A strandof
stele:x, xylem;ph,phloem;
e, endodermis; X37. xylem from each of two poles of the
main root runs into the branch; hence
the lateral roots at the point of attachment are diarch. The
phloem sheath surroundsthe xylemof the branch in the same way
that it does that of the main root; the endodermisalso connects
without a gap. After the branch has separated, the phloem
between the protoxylemand the endodermisdisappears, and the
usual arrangementof the stelar elementsis restored.
In all cases, the branch stele is diarch at the point of attach-
ment,but it may arise froma singlepole of the main root. The
protoxylemstrandsof the branch soon split, and the usual triarch
or tetrarchconditionis established. In one case, a branch arises
froma large tetrarchroot in which significationhas just begun;
each pole is representedby threeor fourtracheidsonly. Two dis-
tinctstrandsof xylem,connectingwith two poles of the main root,
enter the branch. Each xylem strand is surroundedby a sheath
I9141 PETRY-OPHIOGLOSSUM PENDULUM 175

of phloemand endodermis.At a littledistance,the endodermal

sheathscomeintocontact,thenfuseto forma singlesheath; the
phloem shows the same behavior. Xylem parenchymathen
appearsbetweenthe twoprotoxylem strandsand theusual diarch
conditionof thebranchsteleis established.

The bud
The vegetativereproduction of the speciesis accomplished by
buds upon the roots,as in 0. vulgatum.As manyas threebuds
wereobservedupona singleroot; thesecondleafoftheoldestbud
was just appearing,whilethe youngestbud was a mereswelling
about 5 mm.fromtheroottip. It is
almost certainthat the plants of a
colony have all developed froma
singleplant by thismethodof vege-
tative propagation; every rhizome
examinedin whichthe base is intact
showsby its connectionwitha root
that it has developedfroma bud;
fig.5 represents such a case.
RoSTOWZEW (7) has describedthe
developmentof the bud in 0. vul- i m
.
gatum. In the second segment of radial F
character: a, stem tip; ib,
the apical cell of the root a new base of a decayedleaf;r,parent
apical cell arises; this producesall root; XI.
stemtissues. The rootapical cell is
retardedfora time,but finallyresumesgrowth. This resultsin
the formation of a bud withits axis approximately at rightangles
to the parentroot.
The development ofthebudhas beenexaminedin 0. pendulum;
it agreesin all importantpointswiththat of 0. vulgatum.The
retardation of the rootgrowthis usuallyless; the bud axis often
diverges abruptlyfromthat of the root. Two rootsusually
less
developupona bud beforethefirstleafis formed. The detailsof
the vascularconnection betweenthe stemdevelopedfroma bud
and theparentrootwillbe describedlater.
176 BOTANICAL GAZETTE [MARCH

The stem
The largest rhizome of the Australian material measures 0.7
cm. in diameterand 2. 5 cm. in length. There are i8 leaves still
attached and the bases of 5 othersare evident. All the rhizomes,
with the exception of the very youngest,give evidence of having
grownin a horizontalposition; fieldobservationsby Dr. CHAMBER-
LAINconfirmthis. The leaf bases, however are not restrictedto
the dorsal side, as stated by CAMPBELL(3) forthis species,but are
attached in a spiral about the stem. The distributionis irregular;
on young stems the leaves are rathercrowdedand attached along
a spiral with about a 3 arrangement. On older stems the distri-
butionapproximatesa 36arrangement;fig.6 showsthearrangement
of the leaves on the largestspecimen
, secured. The bases of the leaves in-
,',, a- - - ,,,\>\ serted on the lower side curve round
Ii' --'\ , -\ \\'\
the rhizomeand produce the appear-
*@*! ** . 1! ance of a 2-ranked arrangement,as
,1" .in Helminthostachys;but there is
''0\\"\.- --.'/i' nothingin the insertionof the leaves
or the structureof the stem to indi-
cate true dorsiventrality.
FIG. 6. Diagram of leaf ar- The rhizomesof the Samoanma-
ofa horizontally
rangement grow- trial are decidedly largerthan those
ing rhizome.
from Australia; the oldest one
8
secured,with functional leaves and 7 leaf bases, was I . 2 cm. in
diameterand 4.6 cm. in length. Both theirappearance and their
structureindicate a definiteradial arrangement(fig.5). The leaf
insertionis similar to that of the Australian specimens,but the
leaves are less crowded.
All the rhizomesin which the bases are intact give evidence of
having developed frombuds upon roots in the manner described
above. The connectionof the stem stele with that of the root was
examined in I2 specimens; five methods of development of the
stem stele were found.
In a single specimen,a solid strand of xylem, surroundedby
phloemand endodermis,separates at a considerableangle fromone
of the protoxylemstrands-of a tetrarchroot (fig. 7). The endo-
I9I4] PETR I-OPHIOGLOSSUMI PENDULUM I77

dermis soon disappears; before separation fromthe root stele is

complete. a few cells of parenchymaappear near the centerof the
xylem mass (fig. 7, C). The number of these increases and a
definitepith is established (fig.7, D). A small gap opens through
the xylem, and closes again almost at once; the phloem is not
interrupted. The xylem cylinder dilates rapidly and becomes
oval in section; root steles attach at the points of the oval, leaving
large " root gaps" in the cylinder(fig.7, L). A thirdroot attaches
betweenthe two gaps (fig.7, M), producinga small gap; the three
gaps close at about the same level. The firststrand of the trace

DE

F G H

L M j 0
FIG. 7.-Development of stem stele froma single strand; only the xylem is
shown: r, tetrarchstele of parentroot; rt,roottrace; rg,rootgap; Is, firststrandof
firstleaf; XI2.

of the firstleaf separates from the cylinderat a slightlyhigher

level (fig. 7, N). This is essentiallythe course of developmentof
the stem stele of 0. vulgatumfroma bud.
In five cases, two strands of xylem of varyingshape in cross-
sectionseparate fromtwo of the poles of a triarchor tetrarchroot.
Diagrams of sectionsof one of these specimensare given in fig.8.
Each strandat the point of junction with the root stele has phloem
on two faces, or surroundingit; the phloem disappears almost
immediatelyfromthe adjacent faces and becomes restrictedto the
exteriorsides of the strands. After an interval the two strands
178 BOTANICAL GAZETTE [JMARCH

fuse at one margin,forminga half-cylinderof xylem (fig. 8, E).

Two roots are usually given offsomewhat fartherup (fig. 8, I),
and the marginsof the gap come togetherat about the level of the
base of the firstleaf.
In a single instance,threestrandsarise fromthe threepoles of
a triarchroot. These fusejust below thepoint of connectionof the
firstroot to forma large strand of semicircularcross-section. The
gap betweenthe marginscloses just below the level of the firstleaf.
In anotherspecimen,two strandsseparate fromeach of two of the

4 JL M
FIG. 8.-Development of stemstele fromtwo strands; only the xylemis shown:
px,protoxylem strandsofparentroot; ri,roottraces; X i2.

protoxylemstrandsof a tetrarchroot. These fourorganizea stele

in the manner described above; the opening in the side is closed
ratherlate, at the level of the top of the firstleaf gap.
The stem stele is organizedin a distinctlydifferent way in four
cases. A series of sections throughone of the specimens from
below upward is shownby fig.9. In this case, the xylemmass of
one pole of a diarch root begins to enlarge (fig.9, A-C). Aftera
time, medullation of the enlarged xylem strand occurs by the
appearance of parenchymanear the center (fig. 9 D, E, fig. io).
The endodermisdisappears on the side away fromthe unchanged
1914] PETR Y-OPHIOGLOSSUMI PENDULUM 179

pole of the root. At the same level, gaps appear in the sides of the
cylinder(fig. 9, G, H,) and the xylem separates into two strands.
The smallerof these (r, fig.9, H) is the continuationof the original
root,and at once resumesits originaldiarch character. The larger
strand organizes itselfinto the stem stele. The three strands of
the firstleaf separate (fig. 9, I, J), and at a slightlyhigherlevel
(fig.9, J, K) two roots are attached. The stem stele at this point

A B C

D E F

~~~~~~~~~I

FIG. 9.-Development of stemsteleby medullationof rootstele; only the xylem

is shown: r, stele of parentroot; rt,tracesof firstrootsof new stem; it,trace of first
leaf; X 25.

consists of three strands (fig. 9, L), which organize the mature

formof the stele in the usual way. In the otherrhizomesshowing
thismannerof development,the originalroot is triarchor tetrarch;
afterseparation fromthe stem stele. the root stele usually shows
an increased number of protoxylemstrands and the medullated
conditionshown in fig.4.
In all cases, the stem stele soon assumes its usual form,that of
an ectophloicsiphonostelewith more or less overlappingleaf gaps
ISO BOTANICAL GAZETTE [MARCH

(fig. i I). Two gaps usually occur in the same transversesection,

and cause the xylem of the stem to appear as two curved masses
with concave sides facing (fig. ii, G). Often only a single gap
interruptsthe cylinder(fig.i i, A) or the sectionmay show a com-
plete ring of xylem (fig. I 2, C); very rarely,three gaps overlap.
The leaf gaps are circularor oval and usually very large; 2 mm.
in width by 2.5 mm. in height is an average size. The largest
observed measures 3. 2X3.5 mm. In general the stele is more
compact and shows more
definitelyits cylindricalchar-
acter than in 0. vulgatum. It
p > / g very irregularin out-
isa\often
line; the insertionof leaf and
root strandsusually produces
a modificationof shape, as
shown in figs. i 1,I; 12, E;
7, J, K, etc. It is usually
not straight,but slightlybent
at each leaf gap in the direc-
tion away fromthe leaf; this
is undoubtedly due to the
pressure of the young leaf.
In diameterthesteleis usually
a little less than half that of
the rhizome,that is, 3-5 mm.
FIG. io. Detail of E, fig.9: x, xylem; Besides these leaf gaps,
phi, phloem; px, protoxylem; p, paren- openingsdefinitelyrelated to
chyma; e, endodermis; X88.
roots occur in the cylinder
(figs.7, J, L; 8, J, K; B, C,
II, J). These are more conspicuous
in the largerrhizomes,wherea gap occurs above each root; in the
smallerspecimens,gaps occur in connectionwith about half of the
roots. These openings are narrowerin proportionto theirlength
than are the leaf gaps; they range in size from . 2XO.4 mm. to
0.5 X 2 mm. Where a root is attached to the cylinderimmediately
below the point of insertionof a leaf, a gap is almost invariably
produced; the leaf strands attach to the sides of this gap, which
is thereforecontinuouswith the leaf gap. In cases of this kind,
1914] PETRY-OPHIOGLOSSUU PENDULUM

the gap in the cylinderabove the insertionof a rootmightbe inter-

preted as the beginningof the leaf gap; but in very many cases
openings of an exactly similar nature occur when only a root is
involved. These gaps close in the same manner as leaf gaps; for
conveniencetheymay be referredto as root gaps.

AX ~~~~B CIO Di

Ig~ r

@ -~~~~~~~~~S

I ~~K

FIG. iI -Stele of mature rhizome,showingvarious openings in the cylinder;

onlythe xylemis shown: rg,rootgap; rt,roottrace; ig, incidentalgap; ig, leaf gap;
Is, leaf strands; X4.5.

In addition to these,openingsnot related to outgoingstrands

oftenoccur in the cylinder(fig. ii, B-F). These incidental gaps
sometimesoccur at the margin of a leaf gap; a strand separates
fromthe marginof the gap, runsparallel to it fora time,and later
I82 BOTANICAL GAZETTE [MARCH

fuseswith it. At othertimes,a long narrowslit may occur in the

cylinder. Some of the gaps are relativelylarge, measuringo.6 X
0. 7 mm. Apparentgaps, due to a failureof the xylemparenchyma
to signify, sometimesoccur; in these the phloemis not interrupted.
In most of the incidentalgaps, however,the corticalparenchyma
connectswith that of the pith. as in leaf and root gaps.
Lignificationoccurs firstin the layer of tracheidsnearest the
pith; these firstxylemelementsdiffer in no way fromthoseoutside.
They are true tracheidsloftenvery irregularin shape; lobed and
even branched formsoccur. They are relativelyshort,3-6 times
as long as broad; the walls are reticulatelythickened. The ligni-
fication does not begin at definitepoints, but indiscriminately
throughoutthe innerlayer. It proceeds in an outward direction;
in the mature stems the xylem is 5 or 6 tracheidsin thickness.
Occasional irregulardivisionsoccur withinthe procambiumstrand
aftersignificationhas begun. but thereis no true secondarythick-
ening.
The phloem is uniformlya singlelayer of cells; it is separated
fromthe xylem by a layer of parenchyma3-5 cells in thickness.
Thereis no endodermisexceptin the extremebasal regionand at the
pointsof attachmentof roots; in theseinstancesit is a mereexten-
sion of the root endodermis,and is not to be consideredas related
to the stem stele. The phloem abuts directlyupon the cortical
parenchymacomposed of large spherical or ellipsoidal cells with
intercellularspaces; the cells ofthelayersnextthephloemare some-
what smallerthan those fartherout. The walls of all the cortical
cells are secondarilythickenedwith cellulose,as in the innerregion
of the root cortex; the pits are much larger. The pith is in all
respectssimilarto the cortex; the cells of the layersnext the xylem
are considerably smaller than the average. There is no starch
storage in any part of the plant, but fats occur in some quantity
in all parts, especially in the pith and cortex of the stem. The
growthof the stemis by a tetrahedralapical cell, as in otherspecies
of thegenus; its segmentationwas not examined.
The largest rhizome of the Australian material presents an
interestingvariationof thisusual situation. The base has decayed
and it is impossible to say whether the stem originatedfrom a
I9I41 PETRI7-OPHIOGLOSSUUM PENDULUM I83

bud or not. In the lowest part, the cylinderis already definitely

organized and of the full diameter. There are five gaps corre-
sponding to leaves that were no longer attached, and i8 leaves
were stillin position. This is almost certainlythe oldest specimen
secured.
Up to the level of the fourthleaf gap, the stem cylinderagrees
in all details with the usual form,as describedabove; but about
midway of the gap, a strand of xylem separates fromthe margin
of the gap, swingsover to the centerof the opening,broadens,and
by connectionwith the xylem at either side closes the gap. The
gap of the sixthleaf is closed by a similarstrand. The gaps of the
fifthand seventhleaves are closed in a similarmannerby strands
which arise as procambium in the parenchyma of the openings.
At the level of the fourthleaf, a strand separates fromthe inner
surfaceof the cylinderopposite the point of connectionof a root.
This strandrunsin thepith near the middleofthecylinderfora dis-
tance of about 4.5 mm. and finallycloses the gap of the eighthleaf
in the mannerdescribed. In addition to these fivestrands,there
are in this portion of the stem seven others which arise as pro-
cambiumor by separationfromtheinnersurfaceofthe cylinderand
disappear in the same way, withoutbeing in any way related to
leaf gaps. One of these closes an incidentalgap of the cylinder.
From the level of the eighth leaf to that of the twelfth,there
are no medullarystrands,and the stele presentsthe usual appear-
ance. At this point another system of seven strands appears;
their positions and behavior are shown in fig. I2. Between the
seventeenthand twentiethleaves, no medullary strands occur;
but at the level of the latter, three more strands appear as pro-
cambium in the pith. One appears in the opening caused by the
twenty-first leaf and closes that gap; anotherappears in the center
of the base of the twenty-secondleaf,and moves inwardand closes
thegap. The third,arisingnear the centerofthecylinder,branches
once or twice; one branchis recognizableas a procambiumstrand
at a shortdistance fromthe apical region.
In all, 23 medullarystrandsoccur: i6 arise as procambiumin
the pith. 5 as branchesfromthe inner surfaceof the cylinder,and
2 as branches of other strands; 9 of them are concernedin the
I 84 BOTANICAL GAZETTE [MARCH

closure of leaf gaps, 7 fuse with the inner surfaceof the cylinder,
4 disappear by fusionwith other strands,and 3 disappear by the
gradual fadingout of theprocambium. They rangein lengthfrom

-A % a t C a eD

N 0 K

FIG. I 2.-Stele withmedullarystrands; onlythexylemis shown: n;I,firstmedul-

lary strand; X4.5.
19I41 PETRY-OPHIOGLOSSUM PENDULUM i85

3 or 4 tracheidsup to 5.4 mm. in the case of the strandconcerned

in the closure of the gap of the fifteenthleaf, as shown in fig. I 2.
A cross-sectionof one of the strandsis shownin fig. I3; they con-
sist of xylem and parenchyma,without a trace of phloem. The
largest show 30-40 tracheids in cross-section; 8-I 2 tracheids is
the usual size. No protoxylemcan be identified; the tracheids
all resemblethose of the cylinder,

FIG. I3.-Detail ofa medullary

strand;X236

In onlyone otherrhizomewas anythingof thisnatureobserved.

This specimenis fromSamoa, and has only threeleaves; the base
is preservedand shows its originfroma bud. At the level of the
second leaf a small strand,consistingof a fewtracheidsonly,arises
as a procambiumat a shortdistance fromthe innersurfaceof the
cylinder. It verysoon fuseswith the cylinderbut remainsevident
as a ridge for a considerable interval. Fron the point of its
appearance to that of disappearance is about 0.35 mm.
i86 BOTANICAL GAZETTE [MARCH

The leaf
The relationbetweenrootsand leaves is veryvariable. Two
or morerootsusuallyappearon thebuds beforethe formation of
the firstleaf. In the maturestele,rootsoftenconnectwiththe
cylinderimmediately belowa leafgap; but it is equallycommon
to findtwo rootsattachedat the sidesof the base of a leaf. No
definite
relationshipbetweenthetwocan be shown.
As alreadystated,theleavesin exceptional cases reacha length
of 2.7 meters. The stoutcircularor oval petiolemay measure
i cm. or morein diameterand mergesinsensibly into the blade;
thelatterin unbranched leavesis 2-3 cm.in width,butin branched
or lobedleavesmay reach5 or 6 cm.
at the pointof division. Branching
and lobingof the leaf occurs very
commonly in theSamoanmaterial;in
all casestheseparationoccursbeyond
thepointof attachment of thefertile
spike. There is no absciss layer,as
describedforBotrychium virginianum
by JEFFREY (4).
In the bud, the tip of the leafis
curvedover; the fertilespike when
FIG. I4.-Rhizome with firstrecognizableis attached just
spike;xI.
youngleaf:s, fertile beyondthe curve,withits tip directed
towardthe stem. The originalde-
velopmentoftheleafis by an apical cell,but its laterenlargement
is by intercalarygrowth. As indicatedby fig.Iz4,thisbeginsat
the base and proceedstowardthe tip. The portionbetweenthe
spike and the stemmay be 4-5 cm. in lengthwhenthe portion
beyondthe spikeis only 2-3 mm.long. In the matureleaf,the
spikeis attachedat a pointabout one-third of the way fromthe
base to thetipof theleaf.
The numberof strandsseparatingfromthe cylinderto con-
stitutethe leaf trace varies from3 to I2; 5 is the commonest
number. Thereis a generalrelationbetweenthe size of theleaf,
the size of the gap, and the numberof strands. The strands
usuallyattachto the cylinderin a circle,theuppermost one after
I9I4] PETRY-OPHIOGLOSSUM PENDULUM I87

the leafgap is actually closed (fig.12, P; fig.ii, I); in some cases,

however,all the strands connect within the lower portion of the
opening in the cylinder.
The general course of the strandsin the leaf base is at a slight
angle upward fromthe point of connectionwith the cylinder; in
very large leaves, the lower ones may run horizontallyor even in
a downwarddirection. Branchingof some of the strandsusually
occurswithinthe cortexof the stem,so that as many as 20 strands
may be found in the base of the petiole. This may not occur in
small sterileleaves, so that a section of the petiole may show as
fewas 3 strands.
In the sterileleaves, the strands,3-5 in number,arrangethem-
selves in the shape of a C, with the opening directed adaxially.
The strandsin the extremitiesof the arc in passing up throughthe
petiole swing out toward the marginsas the blade is formed,and
all come to lie in a single plane with xylem adaxially directed.
Frequent branchingand anastomosingof the strandsoccur,so that
the numbermay reach I5 or 20 in the blade of the leaf.
In the fertileleaves, the strands,4-I2 in number,are arranged
in a circle or oval as they leave the cylinder,and they maintain
thisarrangementthroughthepetiole. They branchand anastomose
freely,forminga completecylindricalnetwork; no strandsconnect
across the circle. As the petiole broadens and flattens,the strands
arrange themselvesin two series: the outer, consistingof i0-i5
strandswithxylemdirectedadaxially, formthe vascular systemof
the blade; the inner series, of 5-8 strands with xylem abaxially
directed,is reduced by fusionsto 4-6 strandswhich formthe vas-
cular supply of the spike. Anastomosing and branchingoccur
as beforewithin each series, but there is no furtherconnection
betweenthe two systems. Beyond the point of connectionof the
fertilespike, the single seriesof strandsconstitutingthe reticulate
veiningof the blade may increase to as many as 30 in branching
leaves. They forma closed systemwith the exceptionof a very
few small brancheswhich end blindlyin the tip.
In the leaf strandsas theyseparate fromthe cylinder,the xylem
elementsare all alike, but in the base of the petiole and through-
out the leaf the firstformedelementscan be distinguishedby their
I88 BOTANICAL GAZETTE [MARCH

smaller size. These protoxylemelementsare spiral vessels; they

always occur at the inner margin of the xylem, that is, in the
endarch position. The protophloem,consistingof 3 or 4 cells,
develops at the opposite limit of the strand; the later developed
phloemof 6-iO cells is arrangedin 2 or 3 layers. One or two layers
of parenchymaseparate the xylemand phloem in a mature bundle
(fig. I 5).
In the mature strands of the blade, and, to a less extent,of
the petiole, a definite
,fifih bundle sheath is devel-
oped by the thickening
Hi of the walls of 2 or 3
fih- L _ # , layers of parenchyma
surroundingthe vascu-
lar elements (fig. I 5).
The thickeningmaterial-
is celluloseand the walls
are pitted as in similar
tissues of the stem and
root. This bundle
sheath is separatedfrom
7rX the protoxylem by 2 or
3 cells of parenchyma,
but borders directly
of a leaf strand: pph, proto-
FIG. I5.-Detail upon the protophloem.
As a consequence of
phloem; ph, metaphloem; px, protoxylem;X236.
growth within the
bundle sheath, the protophloem in a mature strand is crushed
against and between the cells of the sheath.

The spike
As stated above, the strands of the leaf with xylem directed
abaxially formthe vascular supply of the spike. At the base of
the peduncle these are 4-6 in number; they continue with occa-
sional anastomosingand fusionthroughthe median portionof the
spike. At the base of the fertileportionof the spike, the strands
at the marginof the median region run immediatelyat the base
1914] PETRY-OPHIOGLOSSUM PENDULUM I89

of the sporangia. Fartherup the spike, where the ridge of sterile

tissue which extends into the spore mass is well developed, the
strandoccupies a positionwell withinthisridge(fig.i6, B). There
is no branchingof themarginalstrandbelow the firstfewsporangia;
but between the thirdand fourthsporangia,a shortlateral branch
extendinghalfwayto the edge of the spike usually occurs. Similar
strands occur between all the sporangia above this point (fig. i6,
A). They consistin cross-sectionof io or I2 tracheids,and occupy
the center of the thin wall separating adjacent sporangia. Near
the marginof the spike they spread out in the shape of a fan and
end blindly (fig. i6). The central
strands are reduced by fusion to a -|-
single strand when the tip of the d
spike is reached. Between the ter- - -. '\
minal sporangia this splits into two '
strands which run out above the A
sporangia and end in the fan-shaped
describedabove.
arrangement (jj
The strands of the peduncle and
of the median region of the spike
show the same structureas those of B C
the blade of the leaf,but the bundle FIG. i6. Diagrams of vascu-
sheath is much less developed. In larsystem
ofthespike:A, median
the strands between the sporangia, longitudinalsection;B, transverse
sectionin plane bb; C, transverse
the bundle sheath is absent. The sectioninplaneaa; X4. 5.
protophloem is on the adaxial side
of the xylem,but the later developed phloem elementsmay extend
in the shape of a U about the xylem,whichis always endarch. In
the fan-shaped portion of the strands the phloem cannot be
distinguished; there is no distinctionbetween protoxylemand
metaxylem.
Discussion
The most strikingcharacteristicof the anatomy of this species
is the extremevariabilityof certain structures. The number of
protoxylemstrandsof the root and the numberof strandsconsti-
tutingthe leaf tracevary almost directlywith the size of the organs
concerned. The externalconditionswhich determinewhetherthe
I90 BOTANICAL GAZETTE [MARCH

stemsare large in diameteror small determinein the same way the

number of the strands of the leaf. Such variations may be con-
sidered as produced directlyby growthconditionsand thereforeof
physiologicalinterestonly.
The variabilityin the connectionof the stem stele with that of
the root is related to the positionin which the bud develops; thus,
the bud whichdeveloped the stem stele shownin fig.7 was located
directlyover one of the protoxylemstrandsof the root, while that
of the specimen shown in fig.8 was placed midway between two
such strands. In the same way, the conditionrepresentedin fig.9
may be ascribed to a more gradual divergenceof the rootand stem
axes; fora time,the two apical cellsformeda commontissuewithin
whicha singlestele was developed,as in fig.io. When theangle be-
tween the two axes became greater,separate steles weredeveloped.
It is of course impossibleto say what determinesthese relations
of position of the protoxylemstrands and the stem meristem,or
the rate of divergenceof the two axes; but it seems evident from
these variations that the mannerof developmentof the stem stele
in a bud is controlledby chance or external conditions. Hence
we may conclude that the stelar charactersshown in the develop-
ment frombuds cannot be used in any discussion of phylogeny.
It is to be noted, however,that two featuresare constant; these
are thecollateralarrangementof thestelarelementsand theendarch
position of the protoxylem.
The occurrenceof occasional strands of xylem in the pith is a
featurethat is unique, so far as the writeris aware. The manner
of originof these strands and their behavior suggests somewhat
the medullarystrandsof Marattia; but the absence of phloem dis-
tinguishessharplybetweenthe two cases. In any considerationof
this featureit should be borne in mind that these strandsoccur to
any considerableextentin only a single specimen.
LANG (5) froma study of BotrychiumLunaria has concluded
that the pith of Ophioglossaceae is purely intrastelarin origin.
This opinionis based in part upon the occurrenceof scatteredtra-
cheids in the pith,particularlyin injured rhizomes; this formation
of xylem elementsfromparenchymatouscells of the pith is con-
sideredstrongevidenceof the stelaroriginof thepith. The medul-
I9I4] PETRY-OPHIOGLOSSUM PENDULUM

lary strandsof 0. pendulumafforda much moredefinitecase of this

sort; they are in no way a traumatic response; cells of the pith
develop a procambiumwhich develops into xylemonly; this may
be taken to indicate that all cells of the pith are potentiallyxylem-
producing,and thereforestelar.

Summary
i. The root stele varies from diarch to hexarch. The roots

branch monopodially,and the branches are diarch at the base.

2. Buds develop upon the roots in the same manner as in 0.

vulgatum.
3. The rhizomesare always radial in structure; the leaves are
insertedin an irregularspiral.
4. The connectionof the stem stele with that of the parent root
is very variable; no phylogeneticsignificancecan be attached to
the details of developmentof the stem stele of a bud.
5. The stele of a mature rhizomeis an ectophloic siphonostele
withlarge overlappingleaf gaps; thereis no secondarythickening.
Root gaps usuallyoccurabove thepointsofinsertionofrootstrands.
Incidental gaps not related to outgoingstrandsoccur commonly.
6. In a single large rhizome, numerous xylem strands occur
withinthe pith. These arise fromthe innersurfaceof the cylinder
or as procambium; some of them are concernedin the closure of
leaf gaps, and others disappear as procambiumor by fusionwith
the cylinder. They consistof xylemonly.
7. The vascular supply of the leaf consistsof 3-I2 strands,the
numbervaryingwith the size of the leaf base. These strandsform
a cylindricalnetworkin the petiole; in the lower portion of the
blade, they constitutetwo series of strands with xylem oppositely
directed. The strands with xylem abaxially directed form the
vascular supplyof the spike.

The writeris indebted to ProfessorJOHN M. COULTERformany

suggestionsand criticisms; and to Dr. CHARLES J. CHAMBERLAIN
and Dr. W. J. G. LAND for the material used and for direction
duringthe investigation.
THE UNIVERSITY OF CHICAGO
I92 BOTANICAL GAZETTE [MARCH

LITERATURE CITED
i. BOWER, F. O., Studiesin the morphologyof spore-producingmembers. II.
Ophioglossaceae. London. i896.
2. , Opitioglossumsimplex.Ann. Botany i8:205-2i6. PI. iS. 1904.
3. CAMPBELL, D. H., The Eusporangiatae. CarnegieInst. Pub. I40. Wash-
ington. I 9 I I.
4. JEFFREY, E. C., The gametophyteof Botrychium virginianum.University
TorontoStudies I:I-32. pIs. I-4. i898.
5. LANG, WM. H., Studiesin the morphology and anatomyof the Ophioglos-
saceae, I. On the branchingof Botrychium Lunaria, with notes on the
anatomyof youngand old rhizomes. Ann. Botany 27:203-242. figs. I4.
pIs. 20-21. I9I3.
6. PRANTL, K., Systematischet'bersichtder Ophioglosseen. Ber. Deutsch.
Bot. Gesells. i :348-3 53. i883.
7. RoSTOWZEW, S., Recherchessur l'Ophioglossum vulgatumL. Oversight.
K. D. Vidensk. Selsk. i89i: 54-83. figs.1-17. pis. I, 2.