You are on page 1of 15


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  

Read  more  about  Early  Journal  Content  at­‐jstor/individuals/early-­‐


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  



Three of the species of Pellia-P. epiphylla,P. calycina,and

P. endivaefolia-show morphological differences,especially with
respect to the apical cell, which would suggest a generalized or
possibly an unstable ancestral form. No detailed study of any of
these species has been reported. During the investigationupon
which this account is based, it has been foundthat in P. epiphylla
not onlyare theretransitionsin the methodofgrowth,but also that
the developmentof the antheridiummay followany one of several
It has been generallyaccepted that "the antheridiumof Pellia
is largerthan that of Aneura, but its developmentis very similar,
exceptthat thestalk is multicellular,as it is in otherAnacrogyneae."'
In addition to this method of development,P. epiphylla shows
young antheridiahaving characters of Marchantiales; moreover,
the spermatogenousinitialsmay be cut out in the same way as the
primaryaxial cell of the archegonium.
The divisionof a dorsal cell, the thirdor fourthfromthe apical
cell, by a horizontalcross-wallis the firstevidence of an antheridial
initial. The outerof the two cells formeddivides again, givingthe
three cells of the antheridialrow-the basal cell, the stalk initial,
and the outer cell; the latter by successive divisions gives rise
to the wall cells and spermatogenouscells. Meanwhile, the dorsal
cells, immediatelysurroundingthe antheridialgroup, divide and
become papillate, thereby producing a ring-shaped involucre
(figs. 2, 3). The outer cell of the antheridialgroup divides next.
The position of the wall is significant;if it is vertical and median,
I CAMPBELL, D. H., Mosses and ferns. New York. I905 (p. 92).

Botanical Gazette, vol. 6o] ['34


the successive divisions follow an antheridial sequence; if, how-

ever, the wall is inclined and somewhatremovedfromthe central
position, the cell divisions which follow are similar to those of a
developing archegonium. In the formercase the vertical wall is
followed,by a curved wall on either side, which cuts the vertical
wall as shown in figs.7 and 8. Two similar walls, rotated about
the central axis throughan angle of go' (figs.8, q), complete the
separationof the peripheralregionfromthe centralspermatogenous
region. The firsttwo of these walls may be nearlyparallel at the
base (fig. 7), or they may be at rightangles; similarlythe second
pair. Such an antheridiumis characteristicoftheJungermanniales.
Occasionally the outer cell, mentioned above, is divided into
quadrants by walls at rightangles to the verticalwall (, ii),
in which case four wall cells are cut off by periclinal divisions,
giving also fourspermatogenouscells. The process is similar to
that characteristicof Sphaerocarpusor Marchantiales.
Fig. I4 illustrates the result of a combination of these two
methodsofdevelopment. In fig.I 5 is showna double antheridium;
the two halves have become completelyseparated by the vertical
divisionand each has developed independently. The process may
be compared to the characteristicdevelopment of the double
antheridiaof Anthocerotales.
When the firstwall formedin the outer cell is inclined and
somewhat removed fromthe median position, it is followedby a
second and a third wall, each of which is similarlyplaced, but
revolvedwith respectto each otherthroughan angle of I 26? about
the verticalaxis (figs.I7, i8, 26, 27, 28). A transversewall divides
the central cell into the cap cell and the spermatogenousinitial
(fig. is). The characteristicarchegonialdevelopmentis followed
until the massive spermatogenousgroup begins to be formed(figs.
22, 28), instead of the axial row.
Occasionallythiscriticalthirdwall ofthe antheridiumis inclined
inward instead of outward, as described above (fig. 23). Two
walls similarlyinclined complete the separation of the peripheral
regionfromthe centralspermatogenousinitial (figs.24, 25). This
formis similarto that last describedwiththe exceptionof the incli-
nation of the walls and the resultinglack of the cap cell.

The differencesin these antheridia are emphasized by the

fact that in the firstform(fig. 8) there are two spermatogenous
initials,paired as in Jungermanniales;in the second formthereare
four,arrangedin the formof a quadrant (, as in Marchanti-
ales; and in the thirdand fourththereis but one spermatogenous
initial (figs.I0, 24, 25).
That the structuresdescribedabove are developingantheridia,
and not archegonia,is evidenced by theirpositionand by the pres-
ence of an individual involucre. The antheridia are single and
dorsal,while the archegoniaare groupedin the terminalpocket.

The position of the archegonial group is of considerable mor-
phological importance. CAMPBELL, with referenceto the work of
JANCZEWSKI,states: "The archegonia are formedin groups just
back of the apex but he [JANCZEWSKI]does not seem to have been
able to detect any relationbetween them and the apical cell such
as obtains in Aneura, but it is possible that such a relation does
exist." As mentioned above, the archegonia are terminal and
enclosedin a "pocket," whichis formedby a cup-shapedinvolucral
growth. As will be described more fully, the archegonia arise
froman apical group of cells, any of which may become an arche-
goniuminitial. There is no regularsuccessionin the formationof
archegonia; apparentlyold and young organs are indiscriminately
intermingled. Since the apical group ceases to functionas such
after the productionof archegonia,P. epiphyllamay be regarded
as trulyacrogynous. The involucreis producedby cellswhichare
cut offlaterallyby the apical group (fig.40), and pushed out very
much as the wings during the previous period of growth; in this
event,however,the lateral cells are forcedout on all sides to forma
The structureand developmentof the archegoniumconforms,
in general, to the characteristicliverwortform. Some specific
charactersmay be noted.2 "After the archegonialmothercell is
cut offit does not divide at once by verticalwalls, but a pedicel is
firstcut off [fig. 3A]; after which the upper cell undergoes the
CAMPBELL, D. H., Mosses and ferns. New York. I905 (p. 90).

usual divisions." This characterservesto emphasizethe similarity

which exists between the archegoniumas shown (figs.3I-33), and
the antheridium(figs. i6-2i). Except for the position and pres-
ence or absence of the involucre,these organswould be difficultto
differentiateuntil the spermatogenousgroup or the archegonial
axial row, as the case may be, begins to develop. The cap cell
gives rise to a group consistingof more than the usual number
of cells. The firstdivision of the cap is oftensimultaneouswith
the divisionwhichgives rise to the primaryneck canal cell and the
primaryventralcell (figs.32, 35), and fourcap cells may appear in
cross-sectionwhen thereare but threeneck canal cells and a ventral
cell (fig.36). Fig. 38 shows a group of cap cells, 6 in cross-section,
which may be compared with those which formthe neck of the
archegoniumof Filicineae. JANCZEWSKI reportsthat the number
ofneck canal cells may be as highas i6 or even i8; 9 is the greatest
numberseen by the writer(fig.38). The venterbecomes massive
beforefertilization;it may be 2 or 3 cells in thickness; a many-
celled stalk is also formed(fig.37). In the youngarchegoniumthe
neck has usually only 5 verticalrowsof cells. The cells originating
fromthe thirdwall cell do not divide until the archegoniumap-
proaches maturity(fig. 39).

Methods of growth
A certainformof apical cell may usually be given as character-
isticof a genus or even of a largergroup. In Pellia, however,there
is no such conformity;the apical cell ofP. calycinahas fourcutting
faces,two lateral, a dorsal, and a ventral-the cuneate apical cell.
The dolabrate apical cell of P. endivaefoliahas but two cylindro-
convex cuttingfaces; while that reportedas characteristicof P.
epiphylla,the lenticularcylindricapical cell, has a posteriorconvex
and two lateralcuttingfaces. In the last named species,however,
there are several methods of growth; these cannot be sharply
delimited,but forclearnessfiveratherdistinctformsmay be taken
as characteristicof successive periods of growth.
During the time of intra-capsular gametophytic division a
massive body is formed. There is no regionalgrowth,but all cells
have an equal power of division. This periodof growthis of short

duration. The gametophyte body retains this massive form

until afterthe restingperiod.
The secondperiodofgrowthbeginsby theformationofa cuneate
apical cell. A terminalcell is cleftby a wall inclined about 3Q0
with referenceto the longitudinal axis of the spore; a second
inclined wall bisects the firstat an angle of approximatelygo';
lateral walls complete the cuneate apical cell (fig.4I). The latter
cuts offdorsal and ventral segments (fig. 42) as well as lateral
segments (fig. 43). Such a method of growthis characteristicof
the young gametophyte until about the time of antheridium
The growth of the third period is by means of a lenticular
cylindricapical cell (fig.47). The transitionis somewhatirregular.
Sometimesthe apical cell is more or less equally divided; then the
two halves simultaneouslycut offcells (fig.45) which correspond
to thedorsaland ventralsegmentsoftheprecedingform. The thal-
lus in cross-sectionhas the appearance of being medianallydivided
by a wall. Another transitionformis shown in fig. 44. As the
thallus becomes thickerthe posteriorangle becomes greater,the
two faces being finallyreplaced by one which is curved. The cells
cut offfromtheposteriorface divide rapidly; as manyas 6 segments
may be formedbeforethe next divisionoccurs. The rapid division
of the laterallyplaced cells causes the wings to be protrudedout-
ward and forward(figs. 46, 48, 49). There has been much dis-
cussionregardingthe methodofbranching. HOFMEISTER3 believed
that the centralpapilla ("Mittellappen") shown in fig.I4 was the
seat of the chief apex; hence that there is no true dichotomy.
LEITGEB4 states that the origin of the central papilla is from a
marginal cell. When branching takes place, the apical cell,
instead of cuttingoffa lateral segment,divides equally or almost
so; each of these daughter cells assumes apical characters. The
central papilla is produced by the crowding together of lateral
segments from the two apical cells. It later develops into a
centrallobe correspondingto fusedwings. Branchingis essentially

3 HOFMEISTER, W., HigherCryptogamia. Ray Society. i862.

4 LEITGEB, H., Untersuchungenfiberdie Lebermoose. i882. Vols. II and III.

dichotomous; although the thallus may appear to have a central

axis along the main line of growth,it is a matter of comparative
rapidity in growth rather than origin. This formof apical cell
is continuedthroughoutthe antheridialperiod.
The fourthperiod of growthis terminaland regional; it is con-
cerned with the productionof the archegonialpocket. The cells
surroundingthe original apical cell assume the power of cutting
offlateral and posteriorsegments. The regionof growthis in the
formof a terminaldisk. Lateral segmentsare crowdedout on all
sides to produce the continuous,cup-shaped involucre. A cross-
section in any plane is similar to a horizontalsection throughthe
growingregionat the time of branching(fig.4q). This growthis
checked by the productionof archegonia; any of the surfacecells
of the pocket may produce an archegonium(fig.40). About the
time of fertilizationthe last period of growthbegins.
Any furthergrowthis in connectionwith the developingsporo-
phyte. Starch accumulates in the cells surroundingthe foot and
a massive growth takes place. Usually only one sporophyte
develops in each pocket; a rather thick calyptra is formedabout
it, and sterile archegonia are carried along; these are to be seen
attached to the surface of the calyptra. The firstgrowthof the
gametophyteis massive, similarlythe last.

Relation of antheridiumand archegonium

The relationshipof the various formsof antheridia and their
relationto the archegoniumis demonstratedby the occurrenceof
antheridiain a single species, Pellia epiphylla,which are similarin
developmentto each of these organs. The Marchantiales condi-
tionis generallyto be foundamong thefirstantheridia; theJunger-
mannialesformis most dominant at the middle period, occurring,
however,throughoutthe complete antheridialperiod. The arche-
gonial formis usually found among the last antheridia to be pro-
duced. There is an evident time relation between these forms.
Moreover,one may be regarded as derived fromanother through
a series of progressivesterilizations (text fig. i).5 In the form
5 Cf. HOFMEISTER, The higherCryptogamia. Untersuchungen
tiberdie Leber-
moose. Vol. II (Pellia).

characteristicof Marchantiales (figs.A, B, C), the verticalmedian

wall (I) is followed by two walls (2 and 3) at rightangles to the
former,therebyformingquadrants (A, B, C, D). Four periclinal
walls (4, 5, 6, 7) forma sterilewall cell and a spermatogenouscell


X 3X2 2

1 1 '1



1 ' 1 1

FIG. i.-Explanation in text

fromeach quadrant. In the Jungermannialesform(figs.D, E, F)

the median verticalwall is followedby two walls correspondingto
the second and thirdabove, but somewhatinclined (2, 3, fig. E).
Periclinal walls (5, 7) form two centrallyplaced spermatogenous
cells. The quadrantsA and C have become sterilewall cells; only
the quadrants B and D give rise to spermatogenouscells. In the

archegonialform(G, H, I) the firstwall (i) is not median,but some-

what laterallyplaced and inclined; there is no wall no. 3; the
section correspondingto the quadrants C and D does not divide,
but remains as a sterilewall cell. Wall 2 again cuts offa section
correspondingto quadrant A, which also persists as a sterile wall
cell. A periclinalwall (5) divides the section correspondingto
quadrant B into a wall cell and a spermatogenouscell. Three
of the quadrants have become sterile. Starting with the Mar-
chantiales form of antheridia, by the sterilizationof alternate
quadrants the Jungermannialesform is derived; and by the
sterilizationof threequadrants the archegonialformresults. The
sterilizationsequence corresponds to the time sequence as de-
scribed above.
Only one step is lackingin thisseries to completethe transition
fromtheantheridiumofMarchantialesto the archegonium,namely
the reductionof the spermatogenousmass into a singlerowof cells,
only one of which, the egg, shall be functional. Such transition
forms have been described by DAVIS6 in Marchantiales. The
writerhas seen such " archegonia" with gamete masses in sections
prepared by Dr. W. J. G. LAND. The antheridialformsofP. epi-
phyllafurnish evidencethatthe various typesof antheridiaand the
archegoniumhavehad a commonorigin,possiblygametangiaresem-
bling in structurethe antheridiumof Marchantiales.

The geneticrelationshipsofP. epiphylla,because ofits diversity,
present a rather complex problem. With respect to its relation
to other species of the genus Pellia, if the other species retain
throughoutlife the formof apical cell ascribed to them,it would
follow that P. epiphylla has branched sooner from the general
line of progressand has retained and developed generalizedchar-
acters. As a member of the Jungermanniales,this species is
acrogynousin the sense that growthis checked, apical growthis
stopped,by the productionof archegonia. It differs fromthe more
characteristicAcrogynaein having several regionsof growth,each
of whichis checkedin the same way. It is possible that acrogyny
6 DAVIS, B. M., The originofthe archegonium. Ann.Botany I7: 477-492. I903.

has been reached by several lines of development;hence close rela-

tionship'with "Acrogynae" is not necessarily implied. If the
developmentof the antheridiummay be taken as a basis of classi-
fication,it would seem that Pellia arose from the main line of
advance before the two branches, Jungermannialesand Mar-
chantiales, became separate, even before the archegoniumand
antheridiumbecame definitelydifferentiated in theirmethods of
From Pellia we have evidence regardingthe relation of Jun-
germannialesand Marchantiales. There is in the life history a
transitionfromthe cuneate apical cell of Marchantialesformto the
lenticularcylindricalapical cell foundin certain Jungermanniales;
similarly,the antheridium,already discussed, affordsstrong evi-
dence forthe existenceof such a relation. The evidence tends to
indicatethatMarchantialesare primitiveand thatJungermanniales
are derived.

The antheridium.-The development varies. The dominant
methodis that characteristicof Jungermanniales;formsoccur,not
infrequently, whichare like the antheridiumofMarchantiales,while
othersare like the archegoniumin theirearly development.
The archegonium.-The archegonia are produced fromcells of
the apical group and occur in an archegonial pocket. The diver-
sities from the regular form are few; the large number of neck
canal cells, the extreme development of the cap, the frequent
reductionof the number of tiers of neck wall cells to five, and
the somewhatmassive ventermay be noted. The outercell of the
two resultingfromthe division of the archegonial initial divides
horizontallybeforethe vertical wall is formed.
Methods of growth.-Several periods of growthmay be recog-
nized, each having a specificmethodof growth: the massive; the
period of the cuneate apical cell extending until antheridium
formation; the period of the lenticularcylindricapical cell, or the
antheridialperiod; the period of regional apical growth, or the
period of archegoniumproduction; and the second period of mas-
sive growth,or the period of sporophytedependence.


o O~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
0 :

r,\X - ~00





N 18~~~~~~~
C 0~~



26 28 29o

0 ~ ~ ~ ~ ~ 0

3 03

o~~~~~~~0 32~~~~~~~~~~~3

39~ 000





0 44




I am greatlyindebted to ProfessorJOHN M. COULTER for sug-

gestions and criticisms received, and to Dr. W. J. G. LAND, under
whose direction the work was done.


All drawingsweremade withthe aid of theAbbe cameralucida. Original
magnification of figs.I5, 40, and 50 was 500; of all otherfiguresI730. The
reductionin reproduction is one-third.
FIGS. I-29.-The antheridiumof Pellia epiphylla.
FIGS. i-6.-Longitudinal sectionsof a youngantheridium, suchas is char-
acteristicof Jungermanniales.
FIGS. 7-9.-Cross-sectionsof the same.
FIGS. IO-I2.-Cross-sections of antheridiaof Pellia similarto those of
Sphaerocarpusor Marchantiales.
FIGS. I3, I4.-Sections through the stalk(fig.I3) and spermatogenous
region(fig.IW), the latterillustrating two methodsof growth.
FIG. I 5.-A double antheridium.
FIGS. i6-22.-Longitudinal sectionsof antheridiashowinga development
similarto thatof an archegonium.
FIGS. 23-25.-Similar to last, withthe elimination of the cap cell.
FIGS. 26-29.-Cross-sections of the same.
FIGS. 3o-4o.-The archegonium.
FIGS. 30-38.-Longitudinal sections.
FIG. 39.-Cross-section.
FIG. 40.-Cross-sectionof archegonialpocket.
FIGS. 4I-50.-Growth structures.
FIG. 4I.-A cuneate apical cell in young gametophytewhich has just
escapedfromthe sporecoat.
FIGS. 42, 43.-Sections of sporelingshowingthe segmentationof the
cuneateapical cell; fig.42, vertical; fig.43, horizontalsection.
FIGS. 44, 45.-Transition formsof apical cells.
FIGS. 46-48.-Sections of the lenticularcylindricapical cell, also showing
the regionsurrounding;fig.46, a transverseverticalsection; fig.47, a longi-
tudinal verticalsection; fig. 48, a marginallongitudinalvertical section,
showinghowthelateralsegmentsproducethe wingor lobe.
FIG. 49.-A sectionthrougha regionof apical growth.
FIG. 50.-A youngsporophyteinclosedby a somewhatmassivecalyptra
and the involucre;sterilearchegoniaon calyptra.