John Elihu Hall.

The American journal of science online

. (page 9 of 61)
Online LibraryJohn Elihu HallThe American journal of science → online text (page 9 of 61)
Font size
QR-code for this ebook

Digitized by VjOOQ IC

78 G, R. Wieland — American Fossil Cycads,

tinct, it being held certain that these flowers uniformly bore
disks — were truly bi- or " ampliisporangiate." In the sections
cut from the largest cone, figure 5, the testal walls are clearly
outlined by zonal conservation and one* may plainly see the
more or less collapsed nucellar sacks in the seed interiors.

Notwithstanding this well advanced ovulate growth of the
great majority of the axes of trunk 3, it still seems probable
that amongst the numerous remaining bract groups, various of
which are indicated by the arrows of figure 1, a few may still

Fio. 2 A, Fio. 2 B,


T.J. n. 3. 713*714

Fig. 2. Cycadeoidea Marshiana, [^ x 2-5. B x 1*5.]

A, Longitudinal section through ovalaie cone, fruit No. 3 of Yale trunk
No. 3. As the section finally patsises out a little to one side of the strobilar
axis the terminal brush does not appear. At <S remnant of disk collar.
Compare with figure 7.

B. Longitudinal section of bisporangiate strobilus of Yale trunk No. 3,
composed from two nearly tandem sections, Nos. 713 and 714. The disk has
either wilted loose prematurely or is on the point of dehiscence, haying left
at S precisely the same remnant of the campanular insertion as at <$ in A.

The microsporophyll rachides with short pinnules appear in solid black,
but the decurved rachidal tips do not traverse the central axis, the section
being slightly tangential, as may be exactly noted in figure 3 i4 of the trans-
verse section 761. Compare by number or letter the complementary trans-
verse sections as illustrated in figure 3.

enclose complete flowers. But in view of the results recited
further search was abandoned and the parts of core No. II

Digitized by VjOOQ IC

Further Notes on Floral Sti^uctures.


were cemented together again in their natural place for the
completion of the oriented sections showing the structure of
the entire flower bud, as now to be described.

Fio. 3.


^ ^7 -E c

Fio. 8. Cycadeoidea Marshiana.

The series of transverse sections cnt from the hisporangiate flower bud
No. 2 of Yale trunk No. 3. The exact level of each is indicated by letter
(0y Cy a, by c), and by section namber in the longitudinal view, figure tB.

Observe that the upper figures are more enlarged than the lower ones. At
/ are the fronds, and b b' shows the line where the bract ramentum rests in
that of a leaf base. — U and W mark the trial saw cuts spoken of in the text.

Note that the lower section c (S. 717) traverses the level of disk dehiscence,
while in the two succeeding sections b and a the disk collar is continuous.
Between these two sections b and a the ovulate axis finally terminates, and
between a and C (No. 728) disk division into the fronds/, /occurs.— Con-
tinue to figures^. For exact size compare with figure 2B. — Cf. continua-
tion fig. 3^.

Digitized by VjOOQ IC

80 G, R. Wielaiid — American Fossil Cyeads.

The Flower of Cycadeoidea Marshiana.
I. Yale trunk No. 3.— Figures 2, 3, etc.

The orientation of the series of sections of the sole complete
flower obtained from trunk 3 will at once be apparent on
inspection of the figures. And it will doubtless be granted
that the lesser difficulty of limitation to this single example is
more than compensated for by the resultant fixing of the period
of disk dehiscence at a given immature stage of ovulate growth,
even more accurately than in Cycadeoidea daeotensis. More-
over the series taken as a whole is nearly enough ideal to dis-
play the floral features with precision ; while the traversal of
transverse sections Y17, 726 and 728 by trial saw-cuts and the
necessity of using approximately tandem sections for the longi-
tudinal view figure 2B has in nowise taken away from Mr.
Barkentin's figures. These have all that excellence lent alone
by the study of serial sections by both author and artist with
free use of photographs and joint verification of every detail.

Fig. 3^.

Fio. 3 A. Cycadeoidea Marshiana. Section 761, Yale tmnk No. 3. x 10.

Supplementary figare in cootinaation of transverse series shown in figure
3. The section traverses the microsporophylls at a point just beneath the
down-cnrving of the rachidal apices, the grouping of which plainly appears
at the center of the figure.

In this camera lucida drawing by the writer the ornate sculpturing of the
sporophylls, and especially the grouping and attachment of the synangia
plainly appear.

The longitudinal section 714, figure 2B, stands at right angles to the base
line of the present section, and cross reference to the several figures shows
the location of all the sections.

The silicification of the armor of trunk No. 3 does not
extend to the clear indication of the finer tissues of the
enclosed flowers, although all larger tissue zones and features
are clearly stained and outlined. So that in the bisporangiate
bud one clearly sees the main anatomical details, the peduncle
with its wood zones, the course at least of the bundles given
off, the bracts enveloped deeply in hairy ramentum, the out-
lines of the disk and component fronds with the position of.
the disk and rachidal bundles (figure SA, section 761), the pin-

Digitized by VjOOQ IC

Further Notes on Floral Structures. 81

nules and attached synangia with quite well-marked traces of
wall structure, and tinally the central ovulate cone with its
large pithy axis bearing the young zone of seed stems and
interseminal scales. And we can also see traces of the sporan
gia, either a young condition being indicated, or pollen shed-

FiG. 4.

Fig. 4. Cycadeoidea Mavshiana.

Longitadinal and serial transverse sections through core in which are
embedded two adjacent ovulate cones, x 3/3 nearly. The secant lines in
sections 723-725 indicate the position of the longitudinal section 722. In
the latter the position of the foregoing transverse sections is indicated. At
S in 722 there is a mass of tissue left by the breaking down of a disk.

Digitized by VjOOQ IC

82 G. R, Wieland — American Fossil Cycads.

Fig. 5.

Fig. 5. Cycadeoidea Marshiana.

Two transverse sections in series with those of the succeeding fignre 6.
All five were cut from the core 7 centimeters in diameter drilled from the
larger drill hole noted in figure 1.

The lower section is drawn from a polished surface. It traverses four
peduncles, and is slightly oblique to their axes the upper of which is cut
from 12 to 18""" more proximally than the lower. Continue to fignre 6.

Digitized by VjOOQ IC

Further Notes on Floral Structures. 83

Fio. 6.

^^^ y

Fig. 6. Cycadeoidea Marshiana,

Three sections in series with those of figure 5. The upper transverse sec-
tion No. 756 passes at a distance of 8 to 12°"" above S. 757, figure 5 ; and
the two lower longitudinal sections 726 and 7 lie between S. 757 and S. 756
at exactly right angles to each other as marked. Uach of these sections
traverses the basal part of an ovulate strobilus in which the young seeds
are fairly distinct, and may even approach full size ; though fertilization is
not thought to have occurred.— S. No. 756 and the preceding SS. of fig. 5
are shown x 7/8, and SS. 726, 7 are x 2 nearly.

Digitized by VjOOQ IC

84 G, R, Wieland — American FossU Cycads.

ding and sporangial collapse having occurred. The seeds show
but little structure, being distinctly younger than those of fig-
ure 6, already mentioned as showing testal zones and distinct
nucellar sacks.

Cell walls are however generally obscure : one cannot make
out the bract structure; and similarly the disk and i*achide6,
though very clearly outlined, appear only as an indistinctly
granular groundmass traversed by lighter colored traces of the
bundles, fortunately continuous enough to show the pattern
of the bundle system. But even bo the assemblage or fairly
well-conserved features taken together with the entire outline
of all organs affords a clear view of the form and general
structure of the flower.

On noting that seven rachides are to be seen in sections 726
and 728, and then comparing the series of decurved apices in
section 761, figure 3a, it becomes evident that the disk divides
into twelve small microsporophylls as in tlie young and quite
small flower of Gycadella wyommgeusis (American Fossil
Cycads, figure 93 I) and the very large flowered C. ingens of
the columnar series, instead of dividing into seventeen or
eighteen large staminate fronds as in C. aacotenaia and various
of the WilUamaonia staminate disks or flowere. The point at
which the campanula splits into the separate fronds is accu-
rately located between sections 715 and 728, figure 3, at a
bight of about a centimeter above the apex of the ovulate cone,
which is not a precisely fixable point because ending as a long
thin brush of sterile organs at last almost hair-like.

The length of the microsporophylls can only be estimated
within fairly close limits because of the destruction of the
mid-region of the bud summit by erosion ; but estimating this
loss at about one centimeter and adding for the decurved tips
1*5 centimeters, the full length of the microsporopliylls appears
to have been about 5*5 centimeters. Whence after allowing
for the diameter of the ovulate cone, the flower as imagined
in an arbitrarily expanded form would have a diameter oi ten
or twelve centimeters.

The rachia and pinnules, as one readily sees in the longitu-
dinal section tig. 2B and the transverse section fig. 34, are
much moulded and furrowed by appression faces or even crin-
kled, but withal in a manner producing ornate patterns where
these organs are cut to advantage in regular series. The pin-
nules are broad of base and must tend to confluence with each
other. They are a centimeter long in the mid-rachidal region
and diminish much in length towards both base and apex of
the frond ; so that each frond if laid out flat would have a
more or less elongate-elliptical acuminately tipped, pinnately
parted to pinnately divided form.

Digitized by VjOOQ IC

Further Notes on Floral Structures.


The synangia are well enough advanced in growth to outline
themselves distinctly, being in reality better conserved than
one might expect from the condition of some of the other tis-
sues. Sut the individual sporangia canhot be clearly made
out, and no distinct pollen appears. Inasmuch, therefore, as
the synangia have only from half to two-thirds the size seen in
C, dacotensis buds, in which the size agrees with that of Marat-
tiaceous synangial types, either a somewhat young stage of
growth is indicated, or as is more likely, an incompletely devel-
oped stance due to some failure in floral growth such as would
readily nave been produced by events leading up to fossil-

Fig. 7.



Cycadeoidea Marahiana, Yale tmnk No. 3.

S. 756. Ramentam in transverse section, — study suggested for use in
determining species. The area shown includes the line // separating the
larger leaf -base ramenta from the small bract ramenta. In both cases the
hairs are characteristically one cell thick, except that occasionally the leaf-
base ramental scales are thicker. [Compare with Bennett ites Oibsonianua.]

S. 780. Transverse section through core containing two ovulate cones
which have shed their staminate disks. The section traverses the outer
armor, and cuts through the ovulate cone of the upper axis, but passes above
that of the lower. At I is the saw cut for the longitudinal section 728 shown
in figure 2 A, q. v. Natural size.

On the other hand, the possibility that the synangia like the
flowers were of small size, and the pollen all shed, should not
}ye lost sight of ; and as bearing on this point the supplement-
ary section No. 717 was cut in order to better bring out the fact
that the disk bears the same appearance of wilting and dehis-
cing just above the insertion as in G. dacotensis buds, where an
approach to floral maturity is seemingly indicated.

but in neither case is it necessary to assume that the stam-
inate frond was normally of much larger size than here seen ;
while the ovulate zone is already notably older than in the
C. dacotensis buds, it even being possible that the mature stro-
bilus of trunk 3 did not reach a markedly greater size than in

Digitized by VjOOQ IC

86 G, R. Wieland — American Fossil Cycads.

the largest of the ten axes so far studied. (Cf. fig. 6.) — A fairly
well grown or mature Cddcotensis cone is shown in my Amer-
ican Fossil Cycads, p. 67, under the name C. Marshiana^ and
on the opposite page 66, under the latter name, a form that
belongs to some tnird species not yet satisfactorily determined.
Cone species and the later growth of cones is, however, a long
and difficult subject which cannot be taken up in connection
with discussion of the partly uncertain growth stage of the
flower before us.

II. Flower-Buds of Yale Specimen No. 164.— Figures 8, 9.

This superb silicified cycad was made the subject of special
description with reference to branching in my American Fossil

Fig. 8.

Fig. 8. Cycadeoidea Marshiana. Yale trunk No. 129.

Exceptionally handsome trunk with fruits small and young, but of the
same type as those of figure 9. Minnekahta, South Dakota.

Cycads (structure), pages 41-43, and illustrated in relief on
Plates VII and VIII. Here too, small fruits were supposed
to be young until several thin sections showed the presence of
mature flowers adjudged to be of the same species as those of
trunk No. 3, although sections of the ovulate cones yet require
to be cut, — this task not being relatively urgent since no change
of name is involved in the specific reference of this cycad here
and earlier made.

Digitized by VjOOQ IC

Further Notes on Floral Structures. 87

The first section cut is that of fig. 9 J., showing distinct
agreement of the staininate disk with the corresponding trans-
verse section from the flower of trunk 3 seen in ficrure 3, section
728, and in figure 3 B. And the meaning of the section was
further confirmed by a rigid search all over the surface of the
trunk, taken point by pomt, resulting in the detection of a
single additional example, clear of outline but previously over-
looked. This flower has not been cut from the trunk, where
it appears just as shown in figure 9 B. The synangia are
apparently larger than in trunk 3, which is really yet another
reason for supposing the flower of that trunk to be not quite
fully grown. But caution in judging without thin sections is
required, here or in the case of any flower or strobilus, — more

Fig. 9.

Fia. 9. Cycadeoidea Marshiana.

Two bisporangiate flower buds from trunk 164 of the Yale CoUection, (For
figures of this trunk see American Fossil Cycads, Plate VIII et seq.)

{A) Transverse thin section through the summit of a flower with nine
microsporophylls. The section passes at some distance above the ovulate
cone, and no decurved tips of microsporophylls appear at the center. En-
larged about twice.

{B) Drawing of a portion of the surface of branch (C, Plate VTII, Amer.
Fob. Cycads) showing partly eroded flower bud. The mass of synangia and
pittings corresponding to the rachides of eleven microsporophylls plainly
appear. Shown in natural size.

especially where but a few axes are studied. The number of
disk divisions is clearly eleven ; so that while the study of this
form still rests mainly on macroscopic features, there is little
doubt as to its identity.

Probable Habitus of Cycadeoidea nana of Ward.
Figure 10.

The subject of small cycad flowers and the branching habit
is further illustrated by the quintuply-branched trunk of fig.
10, consisting in a central stem, two large basal and two lesser
lateral brancnes. This exceedingly interesting specimen was

Digitized by VjOOQ IC

88 G, R. Wieland — American Fossil Cycads.

collected by the writer himself while engaged in private explo-
ration carried on in the Black Hills region throughout the
summer of 1902. It was found after some days of patient
search at the well known Minnekahta locality, supposedly ex-
hausted by previous collecting. The specimen has an added
interest and importance as one of the few trunks found in place.
It lay on its side imbedded in a characteristic stratum of stra-
ticulate clayey sandstone, and only a few square inches of the
upper face of the main stqm could be seen.

Had this handsome and finely conserved specimen been
eroded out like most American and the Italian Cycadeoideas,
the basal branches, which were already fractured across their

Fig. 10.

FiQ. 10. Cyeadeoidea nana, x 1/6.

The main central stem gives rise to the two low-set branches I and II,
and the smaller lateral branches III and IV. That numbered IV is the
smallest and fails of the good preservation seen in the others.— Near the
Roman numeral IV is located the ovulate fruit mentioned in the text. —
[Minnekahta, S. D., Author's Collection.]

junction with the main stem, would have become separated,
and probably could never have been brought into their natural
position. It is thus by good fortune that the trunk adds
precious testimony to our knowledge of the branching types.

Moreover, had one of the basal branches been found sepa-
rated it would in all likelihood have been referred to the species
C. nana,* hitherto with some reservation regarded as an un-
* See Ward, loc. cit., pis. clvi, clvii.

Digitized by VjOOQ IC

Further Notes on Floral Structures. S9

branched cycad. To this species therefore the trunk should be
provisionally assigned. Although a rather young form, several
fruits are present, and an ovulate strobilus one centimeter in
diameter and about two in length, has, in strong contrast to the
C. Marshiana strobili figured above, the same flatly convex
type of parenchymatous cushion as C. Wielandi. Whence it
follows that C. daootensis with the allied C. superba^ C. Marsh-
iana^ and C. nanaj include a clear succession of foiTns passing
from the largest of compactly branched cycads to lesser and
finally small-sized and small-flowered, more freely branched

The floral indices of this series are therefore distinct ;
C. daootensis having a larger disk of seventeen or more fronds
and C. Marshiana a much smaller flower with eleven or twelve
fronds, while in C. nana the disk is unknown but the ovulate
cone varies from that of both the foregoing species in its con-
vex instead of elongate parenchymatous cushion. And as will
be illustrated in subsequent studies already fairly complete,
this same transition from the elongate primitive type of stro-
bilar axis to a convex or nearly flat parenchymatous cushion is
as strikingly illustrated by the strictly columnar species of the
Black Hawk localities. There too, an elongate axis is present
in the fruits of the great C, ingens and the tall C, Jenneyana^
while the fine columnar trunk C. excelsa bears the largest cones
with the shortened or reduced axis or cushion type yet seen.

A few of the suggestions arising from the present study of
small cycad flowers m connection with the trunks that bear them
may well be given a tentative record here.

Much stress has been laid in my American Fossil Cycads on
the process of branch formation with increased flower output,
in view of which Cycadeoidea Marshiana and C. nana have
much interest as now shown to typify the extreme of branch devel-
opment and floral reduction in the Oycadeoidem. But it is not
necessary to assume this family to be a direct derivative of
simple stemmed types ; its immediate ancestry may have had
freely branched forms with much slenderer stems. Even with-
out considering Cordaites it is obvious that free branching of
gymnosperms and reduction of floral organs are very ancient.

It is tnen among the Williamson ian tribe that the search for
types of branch development and floral reduction must be con-
tinued. Nor can the role of the threefold process of branching,
sporophyll reduction, and acquisition of the angiospermous
emplacement in older cycadophytans have been a meaningless
one. This course of change must have been in evidence by
Permian time, while Wielandiella with wholly reduced sta-

Am. Jour. Sci.— Foubth Series, Vol. XXXIII, No. 194.— Fxbruabt, 1913.


Digitized by VjOOQ IC


O, R. Wieland — American Fossil Cycads.

mens shows it to have been far advanced in the Trias. It
indeed suggests an evolutionary process that may well have
been the most significant of all that went on amongst the plants
of the early Mesozoic forests.

Even in the Cycadeoidean side line reduction of flowers from

primitive crowns had proceeded
Fio. 11. well beyond the mere limits of

floral size in the Angiosperms,
and the conviction grows that
such changes were the identical
ones responsible for the advent
of these now dominant plants.
Twelve years ago this seemed
the first probability. To-day it
is a theory undergoing demon-
stration, whether we regard the
Angiosperms as monophyletic or,
as seems vastly more reasonable
in the light of the once dominant
cycadophytan plexus, polyphyletic.
Cosmopolitan and plastic races,
already ancient of lineage, are
better conceived of as moving
forward en masse with little loss
in diversity of features.

Regarding the geneial outlines
of such an evolutionary course,
fair inferences were possible with
Cycadeoidea alone well known.
That this type had an ancestry in
which spiral insertion of leafy fer-
tile organs on a main axis was as
distinct as in the female Cycas
was obvious. And that these
organs were of strictly Cycadalean
nature appeared likewise evident
before the restoration of Wei-
trichia^ equally balanced in its
characters between Cycadaceans
and Cycadeoideans, or perchance
even representing one of those
long occulted Medullosans. But
with this form before us we know,
in fine, that the more or less dis-
tinctly monaxial insertion of fer-
tile organs was long retained, that
the whorl of stamens with disk growth early appeared, and
that concomitant floral reduction and branch formation were
at least in part late.

Fig. 11.

Cycadeoidea turrita.
X 25.

Mature seed coDtaining em-
bryo cut in nearly the true
median plane showing the mi-
cropylar tube of three layers,
a heavy inner palisaded, a thin
middle, and a lighter outer pal-
isaded layer, — the tube interior
being filled with a soft tissue.
Two of the five to six wings of
the * * blow off " appear. These
are the smallest mature seeds
of Cycadeoidea so far known.

Digitized by VjOOQ IC

Further Notes on Floral StrucPuree^ 91

Weltrichia, as an early member of* the Williamsonian tribe,
of which the habitns is now so fortunately knowa^Cycadeoidea^
and WielandieUa thus stand in reciprocal relationship. The
first is simple, primitive, cycadean, ont fairly started on the
conrse of change leading towards angiospermons floral types.
The second has by reason of silicification added most to our
knowledge, but its xerophvtic, stereotyped features have long
obscured our conception oi how truly generalized the cycado-
phytans really are in stem leaf and fruit. While the third with
its slender freely dichotomized branches and laminar leaves, in
reality but little removed from any pinnately net-veined type,
has ail but completed the process of microsporophyll reduction.

More or less m contrast to this staminate reduction, the com-
plexly organized Cycadophytan seeds are uniformly ancient of
type. In their coats are best seen Cycadofilicalean or Pterido-
spermous features that add not a little in gaining ideas of race
relationship ; although at the same time both style and stig-
matic surface begin to appear as structures of secondary origin
and function. It is in the seeds that but recently the distinct
resemblances to Gnetum have been noted that may go to indi-
cate this as some long persistent non-plastic form that origi-
nated when the great races preceding tlie Angiosperms began
to convert Paleozoic structures to Mesozoic needs.

And similarly the hiatus between flower and inflorescence,
once held all but impassable, may be but slight if Tumhoa
proves to be another such a laterally related but even older and
more stereotyped line ; it being quite conceivable that con-
tinued reduction of lateral branches bearing, like those of
Cycadeoidea^ flowers derived from strobilar crowns, could
finally give rise to Tumboan and other types of inflorescence.

Digitized by VjOOQ IC

Smith — The Occurrence of Coral Reef 8

Online LibraryJohn Elihu HallThe American journal of science → online text (page 9 of 61)