Full text of "Oogenesis in Saprolegnia"

THE DECENNIAL PUBLICATIONS OF
THE UNIVERSITY OF CHICAGO

THE DECENNIAL PUBLICATIONS

ISSUED IN COMMEMORATION OF THE COMPLETION OF THE FIRST TEN
YEARS OF THE UNIVERSITY'S EXISTENCE

AUTHORIZED BY THE BOARD OF TRUSTEES ON THE RECOMMENDATION
OF THE PRESIDENT AND SENATE

EDITED BY A COMMITTEE APPOINTED BY THE SENATE

EDWARD CAPPS
STARR WILLARD CUTTING ROLLIN D. SALISBURY

JAMES ROWLAND ANGELL WILLIAM I. THOMAS SHAILER MATHEWS

CAUL DARLING BUCK FREDERIC IVES CARPENTER OSKAR BOLZA

JULIUS STIEGLITZ JACQUES LOEB

THESE VOLUMES ARE DEDICATED

TO THE MEN AND WOMEN

OF OUR TIME AND COUNTRY WHO BY WISE AND GENEROUS GIVING

HAVE ENCOURAGED THE SEARCH AFTER TRUTH

IN ALL DEPARTMENTS OF KNOWLEDGE

INVESTIGATIONS

THE UNIVERSITY OF CHICAGO

FOUNDED BY JOHN D. EOCKEFELLEK

INVESTIGATIONS EEPEESENTING
THE DEPAETMENTS

ZOOLOGY ANATOMY PHYSIOLOGY NEUROLOGY
BOTANY PATHOLOGY BACTERIOLOGY

THE DECENNIAL PUBLICATIONS
FIRST SERIES VOLUME X

CHICAGO

THE UNIVERSITY OF CHICAGO PRESS

1903

CONTENTS

I. On the Production and Suppression of Muscular Twitchings and

Hypersensitiveness op the Skin by Electrolytes - - - 1

By Jacques Loeb, Professor and Head of the Department of Physi-
ology

II. On a Formula for Determining the Weight of the Central Ner-
vous System of the Frog from the Weight and Length of
its Entire Body - - - - - - - - -15

By Henry H. Donaldson, Professor and Head of the Department of
Neurology

III. The Development of the Colors and Color Patterns of Coleop-

TERA, WITH Observations upon the Development of Color in
Other Orders of Insects (with Plates I-III) _ . . 31

By William Laweence Tower, Assistant in Embryology

IV. The Artificial Production of Spores in Monas by a Keduction

of the Temperature __- 71

By Arthur W. Greeley, Assistant in Physiology

V. The Self-Purification of Streams ___--_ 79

By Edwin Cakes Jordan, Associate Professor of Bacteriology

VI. The Lecithans: Their Function in the Life of the Cell - 91

By Waldemar Koch, Assistant in Pharmacology

VII. A Contribution to the Physical Analysis of the Phenomena of

Absorption of Liquids by Animal Tissues _ - _ - 103
By Ralph Waldo Webster, Assistant in Physiological Chemistry

VIII. The Distribution of Blood-Vessels in the Labyrinth of the Ear

OF Sus ScROFA Domesticus (with Plates V-XII) . _ - 135

By George E. Shambauqh, Instructor in Anatomy of the Ear, Nose,
and Throat

127055

X Contents

IX. The Animal Ecology op the Cold Spring Sand Spit, with Remabks

ON the Theory of Adaptation 155

By Charles Benedict Davenport, Associate Professor of Zodlogy and
Embryology

X. The Finer Structure of the Neurones in the Nervous System

OF the White Rat (with Plates XIII, XIV) - - - - 177
By Shinkishi Hatai, Research Assistant in Neurology

XI. The Phylogeny of Angiosperms - - 191

By John Merle Coulter, Professor and Head of the Department of
Botany

XII. Studies in Fat Necrosis - - - - -'- - - 197
By H. Gideon Wells, Instructor in Pathology

XIII. Oogenesis in Saprolegnia (with Plates XV, XVI) - - - 225

By Bradley Moore Davis, Assistant Professor of Botany [Hull
Botanical Laboratory]

XIV. The Early Development of Lepidosteus Osseus (with Plates

XVII, XVIII) - - - 259

By Albert Chauncey Eycleshymer, Assistant Professor of Anatomy

XV. The Structure of the Glands of Brunner (with Plates XIX-

XXIV) - - _ _ 277

By Robert Russell Bensley, Assistant Professor of Anatomy

XVI. Mitosis in Pellia (with Plates XXV-XXVII) - - - - 327

By Charles Joseph Chamberlain, Instructor in Morphology and
Cytology

XVII. A Description of the Brains and Spinal Cord of Two Brothers
Dead of Hereditary Ataxia. (Cases XVIII and XX of the
Series in the Family Described by Dr. Sanger Brown); (with
plates XXVIII-XXXIX) -------- 347

By Lewellys Franklin Barker, Professor and Head of the Depart-
ment of Anatomy. With an Introduction by Dr. Sanger Brown

OOGENESIS IN SAPROLEGNIA

Bradley Moore Davis

Although Saprolegnia is a form of considerable interest in connection with the
problem of the so-called multinucleate gametes, nevertheless investigations have not
been carried forward upon it with that attention to cytological detail that has recently
been given to other Phycomycetes, e. g., Albugo, Peronospora, Pythium and Sclero-
spora.

The present paper deals chiefly with the events of oogenesis and a comparison of
this process with the development of zoospores. The material employed was apoga-
mous, indeed apandrous, for specimens were chosen entirely free from antheridia to the
end that the investigation might be relieved from the dispute on the sexuality of these
fungi. However, as will be seen, the results have an important bearing on the well-
known binucleate eggs, assumed by Trow to be stages of fertilization. At the end of
the paper will be found an account, entitled " Theoretical Considerations," which
deals with a number of topics suggested by this study in relation to recent investiga-
tions upon Phycomycetes and Ascomycetes.

The material was isolated in pure cultures and cultivated for several months on
various substrata, during which time the writer had the opportunity of observing and
confirming many of the adaptations recorded by Klebs (1899) in his detailed study of
Saprolegnia mixta. In this period a number of structural peculiarities appeared,
associated with the various sorts of nutrition, and forms arose presenting the charac-
ters of three closely related species, Saprolegnia mixta, S. monoica, and S. fcrax.
The variation was most marked in respect to the presence, absence, or relative quantity
of antheridia which are the most important distinguishing marks of these species.

The original collection bore oogonia with relatively few antheridia {Saprolegnia
mixta), and frequently none. By cultivating the form on a rich substratum — raw
beef or fresh insects — a much more extensive growth of antheridial filaments was
obtained, as in Saprolegnia monoica. On other media — boiled whites and yolks of
eggs and dried beef — the filaments never produced antheridia, but oogonia were
formed abundantly (as in Saprolegnia ferax), normal in size and with numerous
oospores. After three months all cultures ceased to develop antheridia and the num-
ber of oogonia steadily decreased until the cultures reproduced entirely by zoospores.

But it was always possible to get oospores, as Klebs (1899) has shown, by pla-
cing cultures developing zoosporangia under such conditions that the hyphse were no
longer submerged. This may readily be done by removing material from water and
placing it in a dish of cold agar-agar which will furnish enough moisture to support
the fungus for several weeks. The filaments out of water promptly developed oogonia,

227

Oogenesis in Saprolegnia

even when they had the form characteristic of zoosporangia. Such cultures frequently
showed club-shaped oogonia whose eggs were arranged approximtitely in a line.

Chromacetic acid proved to be the most satisfactory fixing agent, but it must be
employed much weaker than the usual formula. One per cent, chromacetic acid
caused immediate contraction of the protoplasm, but a solution one-fourth per cent,
chromic and one-tenth per cent, acetic acid gave excellent results, and presented
advantages of clearness and preservation over weak Fleming, Merkel, corrosive sub-
limate, sublimate acetic, iridium chloride, or picric acid. Paraffin sections were cut
3-5 1^ thick, and generally stained with safranin and gentian violet. The proto-
plasmic structures are so minute as to require lenses of the clearest definition, and the
Zeiss apochromatic objectives 2 mm. and 1.5 mm. with the compensating oculars were
employed throughout the investigation.

OOGENESIS

The accounts of nuclear and cytoplasmic activities in Saprolegnia during oogenesis
present some striking contradictions, and leave untouched some phases of a detailed
but very significant character. Humphrey (1892) was the first author to apply
methods of cytological technique, cutting sections in paraffin, and his studies were
followed by the investigations of Trow (1895, 1899) and Hartog (1895, 1896, 1899).
The last two authors have expressed very divergent views, asserted with a positiveness
that invests their discussions with an atmosphere of personal criticism that need not
be reviewed in this paper. It is necessary, however, to consider certain conclusions
of the earlier authors with which the present writer cannot accord, and it seems best
to do this at the outset, leaving the points of agreement with the present investigation
to be taken up in their proper connections.

It is well known that the oogonium of the Saprolegniales contains many times
more nuclei than the number of eggs ultimately formed. Humphrey and Hartog
believed that the nuclei fuse with one another, thus reducing the sum total until the
requisite number was present. Trow stated that the number was diminished through
degeneration and digestion until it was so small that each egg took but a single
nucleus. The writer has found no evidence of nuclear fusions as reported by Hum-
phrey and Hartog, and in general supports Trow's view of degeneration. However,
there seems to be a reason, not known to Trow, for the selection of the fortunate
nuclei destined to preside over the eggs, and a large part of this paper will deal with
that matter.

It is also well known that the eggs' of the Saprolegniales are not infrequently
binucleate and sometimes trinucleate. Humphrey and Hartog considered such condi-
tions as merely the final stages in that process of general nuclear fusion, the last
pairings whereby the eggs become uninucleate. Trow has made much of these binucleate
eggs, believing the two nuclei to be sexual and one of them introduced by an anther-
idial filament. He has been bold enough to assert sexuality for four members of the

228

Bradley Moore Davis

group: Saprolegnia declina, S. mixta, Achlya Americana, and^. Americana var. cam-
hrica. Nevertheless, Trow presents very little evidence that the so-called "male"
nucleus comes from the antheridial tube, or that the latter structure ever opens into
the eggs. The writer cannot justify Trow's conclusions in this matter, believing them
premature as to evidence and illogical as to probabilities. The present study will
attempt to show that binucleate and trinucleate eggs are to be expected under the
peculiar conditions governing oogenesis.

With respect to cytological details, investigations scattered over so long a period
as twelve years could hardly be expected to agree. Hartog studied from entire
mounts, yet was able to count chromosomes and observe nuclear figures. Trow sec-
tioned in paraffin, and was at first (1895) completely deceived as to the interior
structure of the nucleus and the number of chromosomes. In his second paper, how-
ever, Trow (1899) concedes that the nuclei in the antheridia and oogonia divide
mitotically, but his figures are far from clear as to detail. Trow was also mistaken in
his interpretation of the nucleolus.

The present study will give a more detailed account of nuclear structure and
activities than any previous paper. But the most important contribution relates
to certain cytoplasmic manifestations that seem to determine in large part the results
of oogenesis. These cytoplasmic activities place the process of oogenesis in Sapro-
legnia in new light, bringing it into sympathy with conditions in Albugo, Perono-
spora, and Sclerospora. They are concerned with that cytoplasmic structure termed
the ccenocentrum.

It is not strange that Humphrey, Hartog, and Trow failed to find the ccenocen-
trum, for its recognition demands exceptionally good fixation and staining. It is
probable that Dangeard saw it when he described an oil globule or fatty mass in the
center of the Qgg. It seems possible that Trow may have mistaken it at times for a
centrally placed nucleus, to which it bears a certain resemblance that might make the
two structures indistinguishable in obscurely stained preparations. The ccenocentrum
does not appear until the processes of oogenesis are well under way. Previous to this
period there are nuclear and cytoplasmic activities of considerable import, and they
will be considered first.

It is well known that with the flow of the protoplasm into the swollen tip of a hyplia
there is apparent that peculiar structure of the protoplasm (Plate XV, Fig. 1) significant
of its streaming movement. The nuclei at that time are very small. When the
oogonium is cut off by a septum from the hypha that bears it, the protoplasm
becomes distributed almost homogeneously through the interior (Fig. 2). The nuclei
then increase in size and shortly after show most clearly that detail of structure that
is to be expected in the resting nucleus. This structure agrees with the accounts of
Harper, Waget, Stevens, and myself for the nuclei in other types of fungi, indi-
cating that the conditions among these lower forms are essentially similar to the
nuclear structure of higher plants. As is shown in Figs. 3 and 4 and especially in

229

6 Oogenesis in Saprolegnia

Fig. 6 (Plate XV), there is a nuclear membrane inclosing a well-differentiated
nucleolus, prominent by its size and staining qualities. Much less conspicuous,
but readily demonstrated in well-fixed material, is a loose linin network which con-
tains the chromatic material. Trow's description of a central body containing
cromatin and nucleolar matter, but "neither a nucleolus nor a chromosome," must
have been founded on inferior preparations. There are certainly no complexities
in Saprolegnia comparable to the so-called nucleolus of Spirogyra (Mitzkewitsch,
1898, or Wisselingh, 1900).

There is one mitosis in the oogonium, but previous to that event a number
of vacuoles are developed which generally result in a peripheral arrangement of
the protoplasma around a large central space or vacuole containing cell sap. The
vacuoles begin to appear immediately after the oogonium is cut off from the parent
hypha (Fig. 2). They grow larger and run together as bubbles do in soapsuds
(Fig. 3), until finally there are one or perhaps two large vacuoles in the center, and
occasionally smaller ones near the edge (Plate XV, Figs. 10 and 11). The protoplasm
then lies as a thick peripheral zone, and the nuclei (Fig. 5) are distributed around at
varying distances between the oogonial wall and the boundary of the vacuole.

This is the period when one may expect to find the nuclei in mitosis. The event
happens to most nuclei at about the same time, and good preparations of this stage of
oogenesis are very striking (Fig. 5). The oogonium will be filled with the diamond-
shaped spindles inclosed in nuclear membranes. Three stages of mitosis are shown in
Figs. 7, 8, and 9. It will be noted that the spindle is intranuclear. Fig. 7 presents
the condition just previous to metaphase, with the chromosomes, four in number, at
the nuclear plate and the nucleolus lying outside of the spindle. Fig. 8 is of a stage
shortly after metaphase, when the two sets of daughter-chromosomes have separated
and are about to pass to the poles ; the nucleolus is still present, but smaller and stain-
ing faintly. Fig. 9 is of anaphase, the two groups of daughter-chromosomes, four in
each, lying at the poles of the spindle and the nuclear membrane manifestly about to
disappear. The nucleolus probably dissolves, at least I have never been able to follow it
much beyond metaphase, but surviving, it would of course soon be lost in the granular
cytoplasm after the breaking down of the nuclear membrane. Although granules are
sometimes present at the poles of the spindles, the latter are generally entirely free
from appearances that might suggest centrosomes.

It will be noted that this description of mitosis in Saprolegnia is similar in all
essentials to the accounts of Wager (1896), Stevens (1899, 1901), and myself (1900)
for Albugo; Wager (1900) for Peronospora; Miyake (1901) and Trow (1901) for
Pythium; and Stevens (1902) for Sclerospora. These studies cover a wide range of
forms and material. They agree in describing the spindle as always intranuclear and
without centrosomes. The nucleolus is a structure always distinct from chromatic
material and always, as far as we know, disappearing during mitosis by dissolution or
extrusion into the cytoplasm. The chromosomes are derived from a linin network and

230

Bradley Moore Davis

after mitosis the chromatin returns to the granular condition generally present in
resting nuclei.

Following the mitosis, the oogonium passes into a condition that is exceedingly
difficult to study. The number of nuclei has been doubled by the division, but the
daughter-nuclei are much smaller than the parents. A comparison of Fig. 4 with Fig.
10 will illustrate well the change. It is not the small size, however, that makes the
examination so difficult, but the fact that these nuclei very shortly show signs of
degfeneration. Almost all of the nuclei are affected. The nuclear membrane becomes
indistinct, and its contents finally lie as granular matter in a clear area that resembles,
and probably is, a vacuole. The granular matter is undoubtedly derived in large part
from the nucleolus that fragments, but some of it may be chromatin. The study of the
steps in this process of general degeneration is especially baffling because the progress
is toward a time when the nuclear material becomes indistinguishable from other gran-
ules in the cytoplasm.

It is difficult to understand how Humphrey and Hartog could ever have inter-
preted this process of degeneration as successive nuclear fusions. As Trow pointed
out, successive fusions should give more and more conspicuous nuclei, as the material
accumulated with each union, and consequently an ever-increasing clearness of condi-
tions. In reality, however, we pass from the stage illustrated by Fig. 10 just after mitosis,
to the vague conditions presented in Figs. 11, 14, and 15 (Plate XV). The last two
figures are of oogonia much older than those shown in Figs. 10 and 11, and illustrate
late stages in the process, when the nuclear membranes have mostly disappeared and the
nucleoli and possibly chromatic material lie in vacuoles. Such vacuoles are frequently
elongated, and when they contain two masses of deeply staining material there is sug-
gested a stage in nuclear fusion, and such appearances probably deceived Humphrey
and Hartog. However, the vagueness of structure and manifest waning of the previous
clear definition should have put these observers on their guard. These degenerate
nuclei remain for a long time, even after the eggs are fully formed, and it is quite
impossible to tell with exactness when they lose their structure and functions.

The eggs are formed during the process of nuclear degeneration described above,
and their nuclear structure is really determined by that event. Trow (1899) has given
us a very good account of the general stages in this process of protoplasmic segmenta-
tion, but he did not know the cytological details, and there is reason to believe that he
may have been mistaken in his interpretation of certain structures which he considered
nuclei. The first external indication of protoplasmic segmentation is the gathering of
the contents of the oogonium into denser masses around certain centers, these masses
projecting into the central vacuole and destroying that even outline present in earlier
conditions of the oogonium (Fig. 5). The protoplasm between the egg origins is less
dense, and presently begins to develop small vacuoles (Fig. 12) which run together
until the egg origins are separated by spaces of considerable size (Fig. 13). Many of
these vacuoles break through the films of protoplasm into the central space, which

231

8 Oogenesis in Sapeolegnia

then appears to have put out extensions toward the cell wall. The protoplasm of the
oogonium is exceptionally mobile at this time, and the vacuoles are constantly chan-
ging their forms and positions. In the end the protoplasm gathers more and more
closely around the centers of the spore origins, and finally the latter break away from
one another at all points of mutual contact (Fig. 13), and the several independent pro-
toplasmic masses round themselves off as eggs.

The reader will have noticed in the illustrations of this protoplasmic segmentation
that each egg origin has a deeply stained center surrounded by delicate rays (Figs.
12-15). These star-like structures are very conspicuous under low magnification
(in Figs. 12 and 13, 500 diameters), when the center appears to be a single structure.
In reality it is not a simple unit, but is always composed of at least two structures, a
coenocentrum accompanied by a nucleus. This dual nature is made clear only under
high magnification, with clear preparations of very thin sections. I do not think it
would be possible to understand the structure from entire mounts such as Hartog's.
Hartog probably considered the center as a nucleus alone, and certain of Trow's figures
indicate that he gave a similar interpretation. The coenocentrum is really the key to
many of the problems of oogenesis in Saprolegnia.

The coenocentrum varies in its minute structure with different periods of oogene-
sis. It is at first a small body composed of several granules imbedded in dense
material, from which a number of delicate fibrils radiate into the surrounding cyto-
plasm. The structure stains deeply and resembles an aster. After the eggs are fully
formed the rays disappear and the coenocentrum grows larger, takes on a spherical
form, and resembles a globule of oil or fat. The coenocentrum finally dissolves, some-
times with fragmentation, and completely disappears in the older eggs. The coeno-
centrum is then a structure peculiar to that period of oogenesis characterized by
nuclear degeneration and the segmentation of the protoplasm to form the eggs. It
bears a most important relation to these two events, which are the most difficult to
study in the entire process of oogenesis.

We must begin with the first appearance of the ccenocentra. These structures
may always be found before the differentiation of the egg origins, at the time when
the oogonium is filled with degenerating nuclei. The latter lie scattered through the
cytoplasm, as is shown in Figs. 14 and 15, and exhibit varying degrees of dissolution.
The young ccenocentra are always found in the densest regions of the protoplasm,
portions destined to become egg origins, such as are shown in Figs. 14 and 15. They
are very small at first and would scarcely be noticed except for the radiating fibrils
that mark their position. They increase in size as the egg origins take more definite
form (Fig. 16, Plate XVI).

An examination of Figs. 14, 15, and 16 will show at the side of each coenocen-
trum a small nucleus. This structure is very small at early periods of oogenesis
(Figs. 14 and 15) and scarcely more clear than many of the degenerating nuclei in the
neighborhood. But as oogenesis proceeds the nucleus accompanying the coenocentrum

232

Bradley Mooee Davis 9

grows larger and increases greatly in staining material (Fig. 16). When the eggs are
fully formed this nucleus is many times larger than at the first appearance of the
ccenocentrum, as may be seen by comparing Figs. 17-21 (Plate XVI) with Figs. 14
and 15, which are all magnified 1,000 diameters. One would hardly think it possible
that the large nucleus present in the center of the mature egg was ever so small as the
degenerating nuclei whose remains may be found in advanced stages of oogenesis
(Fig. 16), and sometimes even in the fully formed eggs (Figs. 17 and 23). But there
seems to be no doubt of this. The nucleus destined to preside over the egg is at first
indistinguishable in size or structure from many of its neighbors.

What should lead to its selection as the egg nucleus ? I can see no other explana-
tion but that its position gives it dynamic advantages, enabling it to survive when its
neighbors lack the metabolic conditions necessary for nuclei and consequently must
degenerate. This conceives the oogonium as too richly stocked with nuclei for the
metabolic conditions of oogenesis, and in consequence the field of a struggle of the
parts ("der Kampf der Theile," Roux).

What is the relation of the ccenocentrum to these events? As we have stated,
the ccenocentrum is not a permanent organ either in the oogonium or the egg. It
appears with the first indications of the egg origins and passes away as the eggs grow
older. It is obviously a transitory structure peculiar to the most active periods of
oogenesis. To the writer the ccenocentrum seems to be the morphological expression
of dynamic activities in the oogonium, and especially in the egg origins at the time
when these are difPerentiated. The ccenocentrum has the appearance of being the
focal point in the center of the egg origins of the metabolic conditions peculiar to
oogenesis. And this offers a very plausible explanation of the survival of the nucleus
which lies nearest the ccenocentrum.

The nucleus most fortunate in its position near the ccenocentrum should be greatly
benefited if this is a region of the protoplasm more favorably nourished than other
parts. It is probable that the ccenocentrum even draws toward itself nuclei within a
certain sphere of attraction. Nuclei may be found with a pointed end extended toward
the ccenocentrum (Figs. 16 and 20, Plate XVI). It will be remembered that Stevens
(1901, showed with great clearness for Albugo Candida and A. Tragopogonis that the
nuclei in the immature eggs stretch toward the coenocentra so that their long dimensions
are frequently twice the width. The nuclei of Saprolegnia are too small to present
conspicuous morphological evidence of this character. But we have the fact that the
favored nucleus is almost always pressed against the ccenocentrum which, together
with the appearance of the nuclei and what we know of the events in Albugo, makes
it quite certain that the ccenocentrum exerts a chemotactic influence.

The changes that come over the egg as it matures are illustrated in Figs. 16 to 21
(Plate XVI), which show the usual uninucleate condition of the egg. Binucleate and
trinucleate eggs will be described in the following paragraphs. The two most important
events of maturation are the increase in size of the nucleus and the gradual dissolution

233

10 Oogenesis in Saprolegnia

and final disappearance of the coenocentrum. The growth of the nucleus involves not
only the extent of the space inclosed in the nuclear membrane (Figs. 17-21), but also
means a great increase in the amount of staining material, chromatic and nucleolar.
The latter must be very many times greater in quantity in old eggs than at the beginning
of oogenesis (compare Fig. 16 with Figs. 20 and 21). The coenocentrum decreases in
size until it becomes a very small globule (Fig. 20), or it may split up into several gran-
ules, which soon become lost in an ill-defined mass of denser protoplasm. The coeno-
centrum finally disappears, and the contents of the egg then arrange themselves around
a central vacuole, with the nucleus taking a peripheral position. This is the structure
of the mature egg, and is illustrated in Fig. 21.

We will now consider some conditions that have given rise to much discussion,
namely, the binucleate and trinucleate eggs. They have been found by Humphrey,
Hartog, and Trow, and the present study indicates that they may be expected in any
member of the Saprolegniales. Trow attached much significance to them as evidence
of sexuality, but his conclusions seem to the writer open to much criticism and will be
taken up presently. Figs. 22-5 (Plate XVI) illustrate several conditions that show how
easily an egg may become binucleate. Suppose two nuclei lie near enough to the coeno-
centrum to share about equally the advantages of position. Then it is not likely that
either will give way to the other. Such conditions in a young egg are shown in Fig. 22.
Fig. 25 represents also a pair of nuclei one above the other and both extended toward
the coenocentrum, which was fast breaking down. Fig. 24 is very interesting. In this
instance the coenocentrum is the center of a mass of protoplasm considerably larger
than the average egg. There are two well-developed nuclei, and the form of the cell
suggests the probability that material which ordinarily would have gone into two egg-
origins has been held together in this instance by the influence of an especially large
coenocentrum. An illustration of quite the reverse condition is shown in Fig. 23, and
is remarkable. Here we have presented an egg with two coenocentra, and at the side
of each a nucleus. There is no doubt from the age of the eggs that the two nuclei in
each of these cases are sister nuclei. It is plain that the processes that work for the seg-
mentation of the protoplasm in the oogonium are complex, not all in the influence of the
coenocentrum, nor yet all in the general activities of the cytoplasm.

Give the egg two nuclei with a fair start over their degenerating neighbors, and
they seem to be able to exist side by side, not differing, as far as one may see, from the
nuclei of uninucleate eggs. The two nuclei may lie far apart, as in Fig 26, or so near
together that they touch, as in Figs. 25 and 27. But in no instance — and I have seen
a great many binucleate eggs — have I ever observed them fusing. Trow (1899)
reported an instance of nuclear fusion in the egg, but the writer thinks we are justified
in waiting for confirmations of this observation before attaching to it the importance
given by that author.

Trinucleate eggs are somewhat rare in Saprolegnia mixta. I have seen hardly
more than a dozen, and these were all rather mature examples. I have never beenfor-

231

Bradley Moore Davis 11

tunate enough to find young stages, periods comparable to Fig. 22, 23, or 24 of the
binucleate eggs. The three nuclei may be grouped close together in the egg (Fig. 28),
or may lie quite separate from one another (Fig. 29). There is no evidence that they
fuse. • The rather meager data at hand indicate that when there are three nuclei in an
egg they are individually smaller than the single nucleus in an ordinary egg (compare
Figs. 28 and 29 with Figs. 19-21). This is to be expected, for in general the three
nuclei share between them the metabolic possibilities of about the same amount of pro-
toplasm as is in the uninucleate egg. The trinucleate egg probably develops, as does
the binucleate, from an egg origin in which more than one nucleus by fortunate posi-
tion is able to survive the processes of general degeneration.

Let us now examine Trow's position respecting sexuality in the Saprolegniales.
They are presented most completely in his 1899 paper. I approach this subject with
some diffidence, for it has already been the occasion of detailed discussions of a personal
character (Hartog, 1896, 1899). The matter finally boils down to a question of confi-
dence in Trow's evidence, his account, and his figures. Everyone must admit the pos-
sibility of sexuality in the Saprolegniales, but the question for us is: Does Trow
prove it?

The binucleate egg gave Trow the conviction, as he acknowledges, that fertilization
took place through the introduction of a male nucleus into the egg from an antheridial
tube. But the present studies show that binucleate eggs are quite common in an
undoubted apogamous form, the material being entirely free from antheridial fila-
ments. Moreover, these binucleate eggs have been followed through younger stages
back almost to the period of the egg origins, and we know that these two nuclei were
sisters in the oogonium. To make this point more plain, let the reader contrast the
appearance of the two small nuclei shown in FigS. 22 and 23 with the nuclei in older
eggs (Figs. 19, 20, and 21) and it will be evident that the former have the size and
structure of nuclei in the young oogonium, and not of the fully mature gamete (egg)
nucleus. It should also be noted that Hartog's binucleate eggs were from apogamous
material (Hartog, 1898 and 1899, p. 450) as were also mine.

If, then, apogamous material may have binucleate eggs, and the events of oogene-
sis explain the conditions, we are justified in examining Trow's evidence of sexuality
very critically and demanding of it exceptional fulness and accuracy. We are con-
cerned chiefly with Trow's figures, for they should show most exactly what the investi-
gator really saw. I have been impressed with the lack of detail in many of these figures,
which has led me to think that Trow may have made a number of mistakes which would
quite invalidate his evidence in support of sexuality. Figs. 43, 44, and 46 (Plate XVI)
give appearances labeled "female gameto-nuclei" which are very similar to coenocentra,
and I fear that he was not able to separate these structures in his preparations. Fig. 35
certainly indicates that his material had coenocentra. But the most serious difficulties
are encountered in his drawings of male gameto-nuclei (Figs. 45 and 46). These are
not clear enough to be convincing ; indeed, they seem to the writer to be the remains

235

12 Oogenesis in Saprolegnia

of degenerating nuclei at the periphery of the egg. Side by side with the structures
labeled " male gameto-nuclei " Trow figures bodies very similar in appearance, which
are probably degenerate nuclei. In the face of this uncertainty and seeming con-
tradiction of evidence the illustration of an antheridial filament piercing the egg
(Trow, 1899, Fig. 45) loses much of its weight, and the statement that two nuclei fuse
in the center of the egg (Trow, 1899, Fig. 47) is open to much doubt. The subject
is so difficult that there are abundant opportunities for error, and we are justified in
asking for much more evidence before accepting such important conclusions.

The writer cannot better sum up his attitude toward Trow's opinions on sexuality
in the Saprolegniales than by defining them as not proven and improbable in the face
of the mass of observations upon which botanists have generally agreed that the group
is apogamous. The view of apogamy, formerly resting entirely on the failure to find
antheridial tubes fusing with the eggs, is now supported by the present investigation
on the details of oogenesis. These show that the binucleate egg, formerly difficult
to understand on the theory of apogamy, may arise very naturally in a multi-
nucleate oogonium when the method of oogenesis is as just described for Saprolegnia
mixta.

The binucleate and trinucleate eggs of Saprolegnia are essentially similar to the
multinucleate eggs of Albugo Bliti and A. Portulacae, and the conditions in the young
eggs of A. Candida and A. Tragopogonis, as described by Stevens (1899-1901).
These latter, it will be remembered, contain several potential gamete nuclei, but, so far
as we know, only one of these becomes functional. But it would not be surprising to
find at any time binucleate or trinucleate eggs among species of Albugo that are nor-
mally uninucleate.

In concluding, we must lay emphasis on the importance of the coenocentrum as an
index of the activities peculiar to oogenesis in Albugo, Peronospora, Sclerospora,
Pythium, and Saprolegnia. Although this structure is probably in large part the
expression of activities of the protoplasm as a whole, still there can be no doubt of its
material existence. It is difficult to understand how Trow (1901, p. 291) can question
this point, except that his figures indicate that fine details of structure were not shown
an his preparations.

It would be strange, indeed, if so large a mass of protoplasm as the coenocentnim
should not react in turn on the protoplasm that gave it birth. The coenocentrum is not
a mass of food material, even though much of its granular substance may be the prod-
ucts of metabolism, and the structure as a whole trophoplasmic in character. It is
protoplasm, and as such must be counted a factor in the subtle processes of oogenesis.
Trow's comparison of the coenocentrum to a whirlpool in a river is not good, for there
is unquestionably in this structure the expression of chemical phenomena as well as
physical. The evidence is very strong from Stevens's (1901) work on Albugo, and the
present study on Saprolegnia, that the coenocentrum has a sphere of chemotactic influ-
ence on the nuclei in its neighborhood.

236

Bradley Moore Davis 13

SPOROGENESIS

Except for a recent paper by Timberlake (1902) on Hydrodictyon, we know little
of the details of zoospore formation in either algse or fungi, and the field would cer-
tainly repay investigation. The writer examined the sporangium of Saprolegnia mixta
to contrast the conditions there with the processes of oogenesis, but little came of the
study, the subject not being favorable, except a general confirmation of the accounts of
sporogenesis given by Rothert (1888), Hartog (1888), and Humphrey (1892). If the
oogonium is the homologue of the sporangium, we should expect a general similarity
in the protoplasmic activities of each structure. There is the general agreement that
the protoplasm segments by cleavage planes determined chiefly by vacuoles. But
beyond this the activities of the two structures have little in common and a great many
peculiarities.

As is well known, there is no mitosis in the sporangium. A large number of
nuclei are carried into the tip of the hypha by the accumulation of protoplasm there.
Vacuoles collect and develop in the center of the young sporangium (Fig. 30, Plate
XVI), and, flowing together, form a large central space inclosed in a vacuolar mem-
brane (Fig. 31). The nuclei then lie scattered in the peripheral layer of protoplasm, and
presently clefts appear which work outward between the nuclei from the central vacuole
(Fig. 32). The clefts divide the protoplasm so that it is cut up into polygonal areas,
with clearer regions between. These are the zoospore origins, and each contains a
nucleus.

Rothert's explanations of succeeding conditions, which have also been confirmed
by Humphrey and Hartog, seem entirely satisfactory. The sporangium is in a state of
turgor when the clefts arise and push their way from the central vacuole toward the
periphery. They finally reach the cell wall and immediately make possible the relief
of the fluid in the central vacuole. There is at once a very evident decrease in turgor,
which has an interesting effect on the appearance of the spore origins. The polygonal
areas run together, and the whole sporangium becomes again almost homogeneous in
structure. This means that the contraction of the sporangium brings the spore origins
so close together that the clefts become almost obliterated. The spore origins also
swell. They then begin slowly to separate preliminary to their final rounding off as
zoospores. There is a period when the small masses of protoplasm form a very irregu-
lar network through the sporangium (Fig. 33), and this is followed by a more regular
arrangement (Fig. 31), in which the spore origins are connected by very delicate proto-
plasmic strands. The latter are finally broken and the bodies round off as zoospores.

The writer searched persistently in the sporangium for cytoplasmic centers around
which the process of segmentation might proceed, in the hope that light would be
thrown upon the problem of the coenocentrum, but the examination brought forth no
evidence of siich structures in the sporangium. The nuclei themselves seem to be the
ultimate centers of segmentation. The coenocentrum is then, so far as we know, a
structure peculiar to the oogonium.

237

14 Oogenesis in Sapbolegnia

THEORETICAL CONSIDERATIONS

The writer has once before (Davis, 1900) treated a number of topics suggested by
recent studies on the Phycomycetes. The advances in this field of research, and also
among the Ascomycetes, have been significant, and we seem to be nearing a point where
much clearer conceptions of morphology and phylogeny may result. In this paper we
will take up a number of considerations suggested by this and other investigations
since 1900, and for convenience they will be grouped under headings as follows:

1. Homologies of the coenogamete.

2. Origin and evolution of the coenogamete.

3. Pyronema and coenogametes among the Ascomycetes.

4. Phylogeny of Phycomycetes and Ascomycetes.

5. The nucleus of Phycomycetes in ontogeny.

homologies of the C(EN0GAMETE

The writer suggested the term " coenogamete " (Davis, 1900) as appropriate to
fusing multinucleate masses of protoplasm whose individual nuclei are actually or poten-
tially sexual. Stevens's first paper (1899) on Albugo Bliti really opened the field in
its newer cytological aspects. Since then Harper (1900) has described for the Asco-
mycete Pyronema strikingly similar conditions, as has Juel (1902) for Dipodascus ;
and from the studies of Gruber (1901) we know more about the sexual processes in
the Mucorales. Harper's results will be considered in a special connection. It is
important at the outset that we understand clearly the homologies of the coenogamete.

Are all coenogametes homologous with one another, and from what have they been
derived among the algse ? It will be agreed that the Mucorales, Albugo Bliti, and
Pyronema illustrate completely the conception of a coenogamete. It is part of our
problem to determine the relation of these conditions to the sexual organs in other
species of Albugo, and in Peronospora, Pythium, and the Saprolegniales. There may
be some hesitancy in following the series of homologies that the writer shall propose,
and the evolutionary history to be suggested, but he can see only two possibilities, and
one of these so obscure that it seems almost impossible in the light of our present
knowledge, incomplete as it is.

The most important structures in the coenogamete are the nuclei, and there can
hardly be any question but that they individually stand for energids that among
the algsB are independent uninucleate gametes. Stevens's (1899) term " compound
oosphere" expressed very well this conception of the conditions in Albugo Bliti. It
was employed when this form was the only type known presenting the structure
implied by the phrase, and these conditions might have been purely exceptional.
But we now know from later studies of Stevens (1901) that other species of Albugo
{A. Portulacae, A. Tragopogonis, and A. Candida) have phases of ontogeny identical
with the essential periods of oogenesis in A. Bliti, and may be brought into very
intimate relation to the latter species. We also know that the coenogamete is not

238

Bradley Moore Davis 15

restricted to the Peronosporales, but is characteristic of the Mucorales, and is found
also among the Ascomycetes. It is not likely that we shall retain the phrase " com-
pound oosphere," for a broader conception will probably take its place, but a pur-
pose has been served and a field opened to investigation that was quite undreamed of
by the earlier investigators of the Phycomycetes.

The nuclei of coenogametes are homologous with nuclei in a gametangium destined
to develop independent sexual cells. Hartog's (1891) conception of the nuclei in the
periplasm of Peronospora as representing degenerate gametes has been completely jus-
tified, and there are very good reasons for believing that the nuclear divisions in the
oogonium and antheridium of the Saprolegniales and Peronosporales are "phyloge-
netic reminiscences of the formation of gametes." The attempts to establish special
functions for these mitoses as reduction divisions for the eggs have been inconclusive.

The oogonia and the antheridia of the Peronosporales, Saprolegniales, and Pyro-
nema are the homologues of gametangia, and consequently of that simplest type of
coenogamete, as illustrated in the Mucorales. There is everything in the morphology
of these structures to favor these conclusions, but only recently have we known the
details of protoplasmic organization. When an entire gametangium functions as a
gamete, as in the Mucorales, it becomes a coenogamete. In Pyronema, Albugo, and
the multinucleate eggs of the Saprolegniales the coenogametes are restricted portions
of the protoplasm in such gametangia, but it is obvious that in Pyronema and Albugo
the gametangium behaves as a whole in a manner strictly similar to the fusion of the
coenogametes in the Mucorales. It should be noted that these homologies are quite
independent of the problem of the origin of the coenogametes in the various groups.
That topic will be treated in the next section of the paper.

Stevens (1901) has carried the homologies a step farther in suggesting that the
receptive papilla from the oogonium of the Peronosporales marks the position of the
pore that develops in the gametangia of algsB to give entrance or exit to the sexual ele-
ments. This is a very interesting comparison and is worth following to its limits.
Thus the points of fusion of the coenogametes of a mould may be homologous with the
points of exit of the motile gametes from the gametangium of some algal ancestor.

The term " coenogamete " should be employed in the strict sense indicated when
the term was proposed (Davis, 1900, p. 307). It is a structure containing more than
one gamete nucleus, and generally very many functional or potential gamete nuclei.
It is generally homologous with a gametangium, the binucleate and trinucleate eggs of
the Saprolegniales, and the multinucleate eggs of Sphaeroplea annulina var. Braunii
(Klebahn, 1899) presenting the only exceptions, for the oogonia of Albugo Bliti, A.
Portulacae, and A. Tragopogonis really acts as a whole, and it is hardly possible to
separate in these forms the coenogametes (oospheres) from the gametangia. When we
say that the obgonium of Albugo, Peronospora, Sclerospora, and Pythium acts as a
whole we mean that the periplasm is not to be considered as waste material, but as a
specialized region of the cell, with important functions in relation to the eggs, which

239

16 Oogenesis in Saprolegnia

it helps to protect by assisting in the formation of heavy walls. The Mucorales, Pyro-
nema, and these three species of Albugo furnish the best known illustrations of
coenogametes.

ORIGIN AND EVOLUTION OF THE CCENOGAMETE

There seem to be only two possible sources of the coenogamete. It is conceivable
that a uninucleate sexual element might become multinucleate, perhaps through such
an increase in the protoplasmic content that more than one nucleus would be required
to control satisfactorily its activities. The second possibility is an origin from a mul-
tinucleate gametangium that has given up the production of uninucleate gametes, and
acting as a unit becomes itself a sexual organ, a coenogamete. Such an evolutionary
process would find its analogy in those sporangia (conidia) of certain species of Pyth-
ium and Peronospora, which now germinate as a whole (by a tube) instead of forming
zoospores.

The first possibility has absolutely no evidence in its support. There is no series
of forms whose sexual cells pass from a uninucleate condition to a multinucleate. There
are no indications that such an evolutionary process has ever taken place among plants.
There are only two instances known where eggs, free from periplasm, are multinucle-
ate. The eggs of Albugo are so intimately associated with periplasm that they cannot
be considered apart from the gametangium in which they lie. These two examples are
the binucleate and trinucleate eggs of the Saprolegniales and the multinucleate eggs
of Sphaeroplea annulina var. Braunii. Our investigations of Saprolegnia have shown
that the processes of oogenesis in that group have as an end the sacrifice rather than
the preservation of nuclei, and the uninucleate condition is evidently the goal of evolu-
tion. Klebahn's (1899) and Golenkin's (1899) studies of Sphaeroplea are incomplete in
certain cytological details of oogenesis, and the fact that the eggs of some forms are
uninucleate suggests caution before laying emphasis on the multinucleate condition. It
is possible that further study will relate the multinucleate eggs to the uninucleate, as
in Saprolegnia.

What evidence have we of the second possibility, i. e., the origin of the coeno-
gamete from a multinucleate gametangium which, ceasing to form uninucleate sexual
cells, becomes itself a coenocytic gamete ? Most important is the exceedingly interest-
ing series of four species of Albugo described with so much detail by Stevens (1901).
We cannot take up this investigation except to notice that the four species form a well-
graded series in which the evolutionary direction is clear and very important for the
conclusions that we are striving to establish. The oospheres of Albugo Bliti and A.
Poriulacae contain many functional gamete nuclei, that of A. Tragopogonis several
potential and several functional, and that of A. Candida several potential and one
functional. In this series the coenocentrum is very small in ^. Bliti and A. Poriulacae,
larger in A. Tragopogonis, and very large and strongly chemotactic in A. Candida.
A fifth form has been added to this series by Ruhland (1902), who finds that Albugo
Lepigoni is even more highly specialized than Albugo Candida, since it contains an

210

Bradley Moore Davis 17

extraordinarily large coenocentrum. The evolution in complexity is plainly from A.
Blifi to A. Candida and A. Lepigoni, that is, from the multinucleate egg to the uni-
nucleate. And this series offers the most striking evidence against the evolutionary
possibility considered in the previous paragraph.

Now, the multinucleate eggs of Albugo are not the most primitive types of coeno-
gametes, because they contain only a portion of the total number of nuclei in the
gametangium, many of the sister-nuclei passing into the periplasm. They are not as
simple as the coenogametes of the Mucorales, nor yet as primitive as the oogonium of
Pyronema, which has no periplasm, although it sacrifices a large number of nuclei in
the conjugating tube (trichogyne), and by this specialization presents conditions more
complex than the molds.

It is the specialization of a periplasm simultaneous with the reduction in the
number of functional gamete nuclei that has made possible the elaborately organized
oogonium of Peronospora, Sclerospora, Albugo, Araiospora, and to a lesser degree
Pythium. And the coenocentrum is perhaps most largely responsible for the highest
degree of specialization. The coenocentrum largely influences and perhaps controls
the position and structure of the eggs. The larger the coenocentrum, the more direct
is the effect on neighboring nuclei, and the greater is the benefit to such nuclei as are
fortunate to be within its sphere of operations. So in the struggle for existence
among potential gamete nuclei in the oogonium, the coenocentrum has a power of
assistance that, according to its degree of development, determines the structure of the
egg, whether multinucleate or uninucleate. The evolutionary trend is physiologically
precisely the same as is shown among the algae (Fucales, Vaucheria), when potential
gamete nuclei are sacrificed to provide functional nuclei with a large amount of
richly nourished protoplasm.

But it should be noted that, although the evolutionary processes in the Perono-
sporales have resulted in uninucleate eggs, these structures are not strictly homologous
with the eggs of algae. They are homologous only in the sense that the eggs of Volvox,
Fucus, Vaucheria, Chara, and several other highly developed algae are homologous.
In these algae the eggs have an ancestry from much simpler types of gametes, and
relationships must be traced through these or perhaps through older forms of asexual
spores. The oogonium of the higher Peronosporales has come through a series of
coenogametes of which Albugo Bliti represents a certain stage, but whose earlier forms
must have been simpler. The primitive conditions probably had a structure compar-
able to the coenogametes of the Mucorales, and that type of structure finds its nearest
approach among the algae in the gametangia that discharge numerous gametes, as
illustrated by Cladophora and many of the Siphonales.

But it will immediately be asked: What are we to do with such algal types as
Vaucheria, Sphseroplea, (Edogonium, etc. ? Have they no relation to the fungi ? This
will be considered under the topic "Phylogeny of the Phycomycetes and Ascomycetes."
It is important that we emphasize now the evolutionary process brought out by

241

18 Oogenesis in Saprolegnia

Stevens's work on the four species of Albugo, and extend the results of that study to
the Peronosporales as a whole. Accordingly, we have good reason to believe that the
uninucleate eggs of Albugo Candida, Peronospora, and Pythium have not been derived
from the eggs of algal ancestry, but from coenogametes which passed through the
stage illustrated by Albugo Bliti, and came from much simpler conditions, probably
resembling in many respects the ccenogametes of the moulds and Pyronema.

An origin of the simplest types of coenogametes (moulds and Pyronema) from
gametangia of algae presents certain difficulties that should be discussed. The process
would involve a change in the activities of a structure from one where the nuclei show
a considerable degree of independence to one in which the nuclei co-operate in a
ccenocytic cell that acts as a unit. An evolutionary process comparable to the above
must have taken place with the development of the multinucleate zoospore of Vaucheria
if its nuclei stand for the numerous zoospores generally formed in the terminal spo-
rangia of the Siphonales. And a similar evolution, as has been mentioned before, is
shown in the development among the Peronosporales of conidia (which germinate by
tubes) from sporangia (conidia) that form zoospores. Such conidia and the zoospores
of Vaucheria are not considered the equivalent of tissues, but units in their physio-
logical behavior, just like uninucleate spores.

Similarly, the coenogamete is not the equivalent of a tissue, and must not be
considered as made up of independent gametes associated together because their
cytoplasm is fused into a common mass. It exhibits the same sort of individuality as
any ccenocytic cell or structure. We no longer draw sharp lines between uninucleate
and multinucleate cells, for we realize that the transformation of the first into the
second is a very simple matter, and that the unity of the ccenocyte is not disturbed by
its having several or many nuclei, for these do not occupy fixed positions in the cell,
but wander with the varying movements of the protoplasm. The ccenogamete is as
much an individual cell as the uninucleate gamete, and distinctions can no more be
drawn between these two structures than between the adjacent uninucleate and multi-
nucleate cells of many plants (Chara, Rhodophycese, etc.). In view of its structure
and behavior the term "coenogamete" seems to the writer appropriate.

PYRONEMA AND CCENOGAMETES AMONG THE ASCOMYCETES

Harper's (1900) investigation of Pyronema has established a condition in the
Ascomycetes very similar to that among the Phycomycetes. Pyronema has as con-
spicuous a coenogamete as the Mucorales or Albugo Bliti. Its peculiarities do not
affect the essential cytological structure of the fusing multinucleate masses of proto-
plasm whose gamete nuclei unite in pairs as in Albugo Bliti and probably in the moulds.
The sexual apparatus of Pyronema differs from the moulds chiefly in the development
of that specialized structure the conjugating tube (trichogyne). This organ is mani-
festly of advantage because it affects a union with the antheridium and probably, as
Harper suggests, represents the same sort of outgrowth from a sexual element as a

242

Bradley Moore Davis 19

trichogyne. The nuclei in the conjugating tube break down, and the structure finally
becomes merely the channel through which the protoplasm from the antheridium flows
into the oogonium. This movement of the protoplasm is very similar to Albugo Bliti,
and the resemblance is carried still farther in the distribution and fusion of the gamete
nuclei in pairs throughout the oogonium. The "receptive papilla" of Albugo is
developed from the oogonium, and the conjugating tube in Pyronema may be con-
sidered an elaboration of such a growth tendency. Periplasm is lacking in Pyronema,
but the mass of nucleate protoplasm that passes into the conjugating tube may relieve
the oogonium of those conditions that result in the extensive degeneration of potential
gamete nuclei in Saprolegnia or the somewhat similar conditions affected by the differ-
entiation of a periplasm in the Peronosporales.

It is not to be supposed that the coenogametes of Pyronema are closely related to
those of the Mucorales or the Peronosporales, excepting as all of these structures are the
homologues of gametangia. But it is important that we should recognize this condition
among the Ascomycetes as one that further study may show to be not uncommon in the
group. Juel (1902) reports it for Dipodascus. Miss Nichols's (1896) studies on Cera-
tostoma, while inconclusive in cytological details, are of importance in this connection.
She has described and figured multinucleate oogonia (archicarps) and antheridia,
which are said to fuse. They are apparently coenogametes, and it is probable that
these structures will be found in other genera of the Pyrenomycetes and Discomycetes.
There are several forms whose archicarps suggest a ccenocytic structure (Eremascus,
Ascobolus, Sordaria, Erotium, etc.).

The student of the homologies and evolution of the sexual organs among the
Ascomycetes now finds himself face to face with the same problem that has been pre-
sented to Stevens and myself for the Phycomycetes. What is the relation of the
uninucleate gamete [e. g., Sphserotheca) to the multinucleate? Which condition is the
more primitive?

There is likely to be some confusion of homologies among the sexual organs
of the Ascomycetes. The oogonium (archicarp) of Sphserotheca is morphologically a
gametangium, and so is the antheridium, but both structures are physiologically
gametes. The oogonium and antheridium of Pyronema are morphologically game-
tangia, so that in comparing these two forms we are dealing with homologous struc-
tures. Of course, we use them merely as illustrating certain sexual conditions; indeed,
they are almost the only Ascomycetes whose sexual organs have been thoroughly
studied, with the exception of certain lichens and the Laboulbeniacese, where the con-
ditions are very different and which will be considered later.

The problem then will be: Did the uninucleate condition of the gametangium, as
represented by Sphserotheca, come from a multinucleate gametangium (ccenogamete)
illustrated by Pyronema, or is it the progenitor of the latter? We have no series of
forms in the Ascomycetes such as the four species of Albugo studied by Stevens to
help us to a conclusion. But the problem in the Ascomycetes seems to be identical

243

20 Oogenesis in Saprolegnia

with that of these Phycomycetes discussed in the previous sections of this paper. To
derive a multinucleate gamete (Pyronema) from a uninucleate (Sphserotheca) involves
an evolutionary process quite unknown to botany. To derive a uninucleate gamete
from a coenogamete merely demands a gradual reduction of the number of gamete
nuclei, a process which we know to have taken place in several groups of algae, inde-
pendently of each other, and which is so beautifully shown in Stevens's series of four
species of Albugo.

Harper (1900, pp. 388, 389) seems to be undecided as to the developmental
relation of conditions in Sphaerotheca to such as are presented in Pyronema. He
shows that the oogonium of Sphaerotheca could easily be given the form of Pyronema
by the development of the beak into a conjugating tube with some minor changes in
the position of the antheridium. But he disregards the internal changes necessary to
derive a coenogamete from a uninucleate gamete. And at the end of the same para-
graph he says: "Still I am inclined to believe that the reverse process has taken
place and that the sexual apparatus with the trichogyne represents the more primitive
type for the Ascomycetes." To the writer resemblances of form have very little value
in such comparisons, and relationships must be traced through agreement in the
details of protoplasmic activities. And again, as Harper points out, the general
morphology of the Erysipheae is much higher than that of Pyronema. But Professor
Harper by his last statement has warded off criticism, and perhaps, with the evidence
from Albugo and Saprolegnia before him, he will feel more certain, with the writer,
that the coenogamete when related to the uninucleate gamete always represents more
primitive conditions.

And this conception has a very interesting relation to the possibility of deriving
the trichogynes of lichens and the Laboulbeniaceae from a primitive type of sexual
organ that may have been a coenogamete. Of course, there is no reason why a
uninucleate gamete (archicarp) among the fungi might not develop a simple trichogyne,
as has been done in the Rhodophyceae, but the trichogynes of the lichens and the
Laboulbeniaceae are generally systems of cells quite distinct from the female gamete
(carpogenic cell). These conditions are nowhere presented in the red algae, and it is
very difficult to understand how a uninucleate gamete could develop such elaborate
structures. But taking the suggestion of Harper (1900) that the conjugating tube of
Pyronema is an outgrowth similar to a trichogyne, there are presented possibilities of
various elaborate structural developments, because the outgrowth has so much proto-
plasm and many nuclei to draw upon. The evolutionary tendency of a coenogamete is
to reduce the number of functional gamete nuclei, generally by the sacrifice of many,
but these with accompanying cytoplasm are sometimes employed to advantage in
developing structural adaptations. In the Peronosporales the advantage lies in the
activities displayed by the periplasm in assisting to form the spore wall. Araiospora
(Thaxter, 1896) utilizes the periplasm to develop a cellular envelope surrounding the
egg. The conjugating tube of Pyronema is evidently a desirable specialization,

Bradley Moore Davis 21

insuring a union with the male organ. Perhaps the elaborate multicellular trichogyne
is the result of similar activities on the part of archicarps that are or were cceno-
gametes.

It is obvious that this possibility has very important relations to comparisons that
have been made between the trichogynes of the Ascomycetes and. those of the
Rhodophycese. It is not easy to homologize these structures and it is difficult to
conceive the evolution of any group of the Ascomycetes from the red algge. The
Laboulbeniacese exhibit certain strong resemblances in a general similarity of cell
structure, but peculiarities confront one whenever the comparison is carried into
details. Nevertheless a relation of this group to the Rhodophyceae remains a
possibility, although it can hardly be more than mere speculation until we have much
greater cytological knowledge of sexual processes here and in other Ascomycetes.

But the coenogamete may be found widespread among the Ascomycetes which
suggests a new point of view that is worth attention. It is possible that the coeno-
gamete may become recognized as a primitive type of sexual organ in the Ascomycetes,
as the writer believes it to be for certain regions of the Phycomycetes (Mucorales,
Saprolegniales, and Peronosporales). Perhaps the complex conditions of such highly
specialized groups as the ErysipheaB, lichens, and Laboulbeniacese may be related to
the peculiar activities and possibilities of diverse development in this interesting
sexual cell, the coenogamete. Sphserotheca may readily stand as the last step in a
process of nuclear reduction. Pyronema certainly exhibits the tendency to utilize
superfluous nuclei and protoplasm in developing that advantageous structure the
conjugating tube. And possibly such tendencies might result in the production of
the elaborate trichogynes of the lichens and Laboulbeniacese and in the latter group
the structure that resembles the procarp of the red algse.

PHYLOGENY OF THE PHYCOMYCETES AND ASCOMYCETES

The reader of this paper has probably already noted that some standpoints have
been taken at variance with the generally accepted ideas of relationships among the
Phycomycetes and Ascomycetes, and of these groups to an algal ancestry. A protest
is sure to be offered against the disregard of certain Phycomycetes and algse in the
attempt to derive the Mucorales, Saprolegniales, and Peronosporales from an ancestry
with coenogametes.

There are certain Phycomycetes much closer to the algae than any of the groups
mentioned above. Monoblepharis and Myrioblepharis (Thaxter, 1895) exhibit sexual
organs, zoospores, and vegetative structure with striking resemblances in various
particulars to such algae as Vaucheria, (Edogonium, and Sphaeroplea. The homologies
can hardly be questioned and will not be elaborated here. These fungi, and possibly
some of the Leptomitacese, seem to be close to heterogamous (oosporic) algae and may well
have come from that region of the Thallophytes. The family Leptomitacese includes
some very remarkable types which have been well described by Thaxter (1896). Their

245

22 Oogenesis in Sapbolegnia

position must remain somewhat uncertain until we know the nuclear structure of the
sexual organs, but the general morphology of some forms indicates a relationship to
the Peronosporales. Araiospora (Thaxter, 1896) is likely to prove especially interesting
as illustrating an activity of the periplasm, in forming a cellular envelope around the
oospore, that is not shown in any other type and which has important bearings on
the possibilities of the coenogamete to develop tissues of considerable complexity.

But many difficulties present themselves when the Monoblepharidae are made a
starting-point for a line of ascent to the Peronosporales, as is done by Trow (1901, pp.
306, 307) when he arranges a series Monoblepharis, Saprolegnia, Pythium, and
Albugo (Cystopus). These forms are not so similar that close relationships are
manifest either through morphology or ontogeny. The most favorable interpretation
must grant that they are at present widely divergent and highly specialized types,
even assuming that ancestral forms now extinct might have had more general
characters. Such speculations are, of course, entirely justifiable, if they do no violence
to developmental processes.

However, as has been shown, such an evolution must assume either that uninucleate
gametes became multinucleate or that difiPerentiated eggs (Monoblepharis) lost their
high state of specialization and finally their entire individuality in the coenogamete of
the Peronosporales. Both processes are opposed to what we know has been the
evolutionary history of sexual cells in several divergent and independent groups of
algae. We are called upon to accept a "subjective phylogeny" opposed to well-
established cytological processes.

The situation is somewhat similar to that presented to the Brefeldian school with
respect to the origin of the ascus from the sporangium of a mould. Harper has shown
that the protoplasmic activities of sporogenesis in the sporangium and ascus are along
entirely different lines with nothing in common. To the writer such differences in
cytological processes completely outweigh conclusions from any series of types
presented on a basis of general form resemblance. Form resemblance between the
ascus and sporangium can have very little morphological value until it be accompanied
by evidence satisfactorily explaining the differences of protoplasmic organization and
behavior. And the elaborate phylogenetic structure built by Brefeld and his followers
is sadly in need of a foundation, if not already a ruin. Form resemblance must be in
complete sympathy with cytological conditions to have weight.

Trow (1901) has criticised a developmental line that the writer indicated in 1900,
which, he states, is an attempt to derive Oomycetes from a Zygomycete-like ancestry
and which he considers an example of "subjective phylogeny." I have carefully
examined what was written in that paper (Davis, 1900, pp. 304-9), and, not finding
any reference to specific phylogenetic ancestry, am compelled to suggest to Trow a
more careful reading and citation of that article. I presented there suggestions for
the developmental history of the sexual conditions in the Peronosporales from cceno-
gametes derived from the gametangia of algae. These ccenogametes at a certain stage

246

Bradley Mooee Davis 23

in the process of sexual differentiation would be similar to the sexual organs of the
Mucorales. The moulds were used to illustrate a well-defined sexual condition, which
is not at all suggesting that they are the ancestry of the Peronosporales (Oomycetes).

But the present investigation of Saprolegnia, together with Stevens's (1901) later
studies on Albugo, have strengthened my faith in the suggestions of that former
paper (Davis, 1900). The Mucorales, Saprolegniales, and Peronosporales are
generally acknowledged to be closely related groups, but it seems probable that the
affinities are only through the somewhat similar conditions of sexual organs derived
from the ccenogametes of some common ancestry. There are many peculiarities of
life-habits, life -histories, and methods of asexual reproduction. Of these three groups
the Mucorales presents the simplest conditions of sexuality and illustrates most nearly
the structure of the primitive ccenogamete. The Peronosporales and Saprolegniales
are difficult to relate to one another, for the higher development of the ccenogamete
is apparently progressing along divergent lines. In the Peronosporales the protoplasmic
differentiation in the oogonium determines a centrally placed egg in an enveloping
periplasm, for a single coenocentrum dominates the process of oogenesis. In the
Saprolegniales the ooplasm gathers by cleavage around a number of ccenocentra, and
all the protoplasm passes into the resulting eggs. To the writer the second process
seems less highly specialized than the first and the Saprolegniales lower than the
Peronosporales with respect to sexual processes. But oogenesis in these two groups
shows such marked differences in their evolutionary tendencies that the question of
the relative level of each process has very little import.

The Saprolegniales are more difficult to understand in relation to a ccenogamete
ancestry than the Peronosporales, because the many eggs without periplasm suggest at
once the stage in heterogamy illustrated by Sphseroplea. However, the processes of
oogenesis are probably very different in the two types. The egg origins of Saprolegnia
have a great many potential gamete nuclei, and that stage indicates strongly the
ccenogamete ancestry. By numerical reduction of the gamete nuclei the egg of Sap-
rolegnia has proceeded to a point where it has almost ceased to be a ccenogamete, that
condition only being presented in the bi- and tri-nucleate eggs.

It will be difficult for many to give up the idea that Vaucheria is not a suitable
starting-point for the line of higher Phycomycetes. The chief objection is the incom-
patibility of the processes of oogenesis where a relationship demands agreement even
in the details of cytology. We have only the accounts of Oltmanns (1895), Behrens
(1890), and Klebahn (1892), which are not in complete agreement on some
important points, and perhaps further study may reveal conditions that are only sus-
pected. In considerations of this sort it is important to know the relation that
Vaucheria bears to the algae as a whole. Although generally classed among the
Siphonales, Vaucheria has little in common with that group excepting the coenocytic
thallus. It stands alone as the only heterogamous form (oosporic) in a very large
assemblage characteristically isogamous. Generally taught as a type of the Siphonales,

247

24 Oogenesis in Saprolegnia

Vaucheria is not really representative of that group, which is much better illustrated
by such forms as Codium, Bryopsis, or Penicillus. The affinities of Cladophora with
the Siphonales are now better understood, and we see that this form, except for the
septate thallus — whose cells are, however, multinucleate — has all the characters of the
Siphonales. It is this region of the algae (Cladophora, Codium, etc.) that presents to
the writer's mind conditions most nearly like the ancestry of the Mucorales, Sap-
rolegniales, and Peronosporales, that is, an ancestry whose sexual organs were
coenogametes. However, perhaps, farther studies on the oogonium of Vaucheria may
bring this structure into sympathy with coenogametes.

The coenogamete among the fungi must have come through the homologous struc-
ture among the algae, the gametangium. We cannot suppose that such gametangia
were highly specialized. It is hardly possible that they were heterogamous, for a
highly differentiated oogonium would not be likely to return to conditions as simple as
the primitive coenogamete. The gametangia of such insogamous algae as Cladophora
and Codium present most nearly the structure demanded of the progenitors of the
primitive coenogamete, but, of course, these forms are mentioned only as illustrations
of conditions undoubtedly present in many groups of algae at various periods in their
evolutionary history.

We can only speculate as to the manner in which a gametangium might become
a coenogamete. The writer has already offered some suggestions on this point (Davis,
1900, p. 308), and he is more inclined to them since the recent studies of Harper
(1900) and Stevens (1901) and the present investigation of Saprolegnia. We can
readily conceive the derivation from isogamous algae of groups of aquatic fungi with ter-
minal sporangia discharging motile gametes after the manner of Cladophora. Should
such fungi leave the water and adopt a terrestrial life either as saprophytes (Mucorales)
or parasites (Peronosporales), certain changes in the sexual processes would be very
likely to result. The gametangia could not form and discharge motile gametes except-
ing when wet, and would be compelled to adapt themselves to the aerial environment.
They would be very likely to develop such unity of structure and behavior as is dis-
played in Pythium and Peronospora by those sporangia (conidia) which have given up
the habit of forming zoospores and now germinate by a tube. The gametangium
would become a coenocytic unit with the chemotactic qualities and possibilities
associated with sexuality. These chemotactic influences might be satisfied by the
fusion of the gametangia (coenogametes) in pairs whereby the gamete nuclei would be
able to unite two by two in a common protoplasmic medium. This process would take
the place of the conjugation of motile gametes in water, and apparently satisfy all the
hereditary demands as far as nuclei are concerned. The structure resulting from the
fusion of these simplest coenogametes would be very similar to the zygospore of the
moulds.

Although there are no coenogametes among the algae, the sexual processes in the
Conjugales have some features worth noting in this connection. In the desmids the

248

Bradley Mooee Davis 25

gametes slip from the parent cells and fuse as naked masses of protoplasm. But in
the filamentous forms ZygnemacesB and Mesocarpaceas the energids (gametes) remain
in the respective parent cells which push out conjugating processes. The conjugating
processes are surrounded by a cell wall so they are in every respect similar to the con-
jugating tube of Pyronema or the receptive papilla of the Peronosporales, excepting
that they emanate from a uninucleate cell instead of a ccenocyte. It is important to
note that such conjugation processes in the Phycomycetes and Ascomycetes have their
analogies in the algae, for it might be suggested that the development of such a struc-
ture by a ccenogamete would be difficult. On the contrary, it seems the natural expres-
sion of any cell, whether uninucleate or multinucleate, with chemotactic tendencies to
fuse with its neighbors. It is very probable that the development of such a conjugating
tube in a ccenogamete would be at the point where formerly the naked motile gametes
were discharged, for that place is evidently the seat of important cytoplasmic activities.

To sum up our conception of the Phycomycetes, we must regard them as a group
of several independent phyla. The Chytridiales in morphology and life-histories are
the lowest and resemble the algae at the level of the Protococcales. Monoblepharis,
and probably several other isolated genera, seem most closely related to heterogamous
algae. The Entomophthorales are too highly specialized to be easily derived directly
from algal ancestry and need not be considered in this paper. There are left the
most conspicuous of the Phycomycetes in three orders that agree primarily in having
either typical coenogametes (Mucorales) or sexual organs probably derived from cceno-
gametes (Peronosporales and Saprolegniales).

These three orders can, however, only be related to one another through a com-
mon ancestry whose sexual organs were coenogametes. The Mucorales illustrate most
completely the primitive ccenogamete, and for this reason in part may be considered
rather the lower of the three groups. In the Peronosporales we have an ascending
series from forms such as Albugo Bliti and A. Portulacae with true coenogametes,
although more highly specialized than those of the moulds, to the conditions in A Ibugo
Candida, Peronospora, and Pythium. This advance is evidently such an evolution as
would provide a single uninucleate egg with the richest supply of food and best pro-
tective walls possible. In the Saprolegniales the evolutionary trend is similar in that
a great many potential gamete nuclei are sacrificed to give a uninuclete egg, but we
are not yet prepared to trace exactly the steps in the origin of this oogonium. How-
ever, the probabilities are that it, too, has come from a ccenogamete, and that the seg-
mentation of this protoplasm to form many eggs does not imply a derivation from
heterogamous ancestry, but special peculiarities associated perhaps with the presence
of several coenocentra. Oogenesis in Saprolegnia certainly indicates an ancestry with
coenogametes. The Mucorales, Peronosporales, and Saprolegniales then probably
come from a somewhat similar ancestry with coenogametes, which necessitates their
derivation from isogamous algae at about such a level as is illustrated today by Clado-
phora and forms of the isogamous Siphonales.

249

26 Oogenesis in Saprolegnia

We do not propose to discuss the phylogeny of the Ascomycetes further than to
present its problem with respect to the coenogametes. The difficulty of relating the
diverse sexual organs represented by Sphaerotheca, Pyronema, the lichens, and the
Laboulbeniaceae has led to suggestions that the Ascomycetes are polyphyletic. But
this view has many objections in the essential unity of the ascocarps and general rhythm
of the life-histories throughout the group. Nevertheless, the various types of sexual
reproduction seem very diverse when compared with one another and with conditions
in the algae and other fungi.

However, should the coenogamete be established as a primitive type of sexual organ
here as in the Phycomycetes certain difficulties will be removed. The oogonium (archi-
carp) would be considered a development from the coenogamete along a well-estab-
lished evolutionary line, that of numerical nuclear reduction. The evolutionary trend
of the coenogamete would then be toward the uninucleate oogonium (Sphserotheca) fol-
lowing the tendency of sexual evolution so well recognized in the algse. The groups
of the Discomycetes and Pyrenomycetes would then readily arrange themselves accord-
ing to the structure of the ascocarp and general vegetative complexity.

There would be left the lichens and Laboulbeniaceae, whose trichogynes at least
suggest the Rhodophyceae, while in the latter group there are certain histological resem-
blances to this same group of algae. Granting these possible affinities, it is neverthe-
less very difficult to conceive the multicellular trichogyne as derived from the simple
structure of the red algae. It must also be borne in mind that the structure of the
ascocarp, especially among the lichens, gives no suggestion of a cystocarp, but, on the
contrary, presents a structure identical with the fructification of other Ascomycetes.
Were it possible for the coenogamete to develop a multicellular trichogyne (there is a
multinucleate one in Pyronema), then evolutionary lines might be established that
would lead very naturally into the lichens and Laboulbeniaceae. Such trichogynes
would be another form of expression of this remarkable structure, the coenogamete,
which is able to utilize superfluous protoplasm in such a variety of ways.

In this connection it is interesting to sum up the various ways in which the super-
fluous protoplasm of a coenogamete may assert itself. It may form a periplasm of
importance in developing the spore wall (Peronosporales). It may form a surrounding
tissue from such periplasm (Araiospora). It may develop a conjugating tube (Pyro-
nema). And finally we suggest the possibility of multicellular trichogynes derived from
coenogametes. While this cannot be more than a speculation, nevertheless cytological
and developmental investigations among the lichens and Laboulbeniaceae in relation to
these possibilities are sure to bring forth interesting results.

We may then conceive the Ascomycetes as presenting two important evolutionary
lines derived from a primitive coenocytic type of sexual organ (coenogamete). The first,
through numerical reduction of potential gamete nuclei, results in uninucleate sexual
organs (Sphaerotheca). The second line supposes the utilization of such potential
gamete nuclei with cytoplasm to develop such secondary sexual structures as the

250

Bradley Moore Davis 27

conjugating tube of Pyronema and the trichogynes and procarpic apparatus of the
lichens and Laboulbeniacese.

It is difficult to relate the account of Juel (1902) for Dipodascus to conditions
in other ccenogametes. Juel believes that there is but one sexual nucleus in each of
these multinucleate gametes, the others being " vegetative " : that there is only one
fusion nucleus in the fusion cell. This gives rise to a series of nuclei around which
the spores develop in the sac and the "vegetative" nuclei degenerate. The details of
the nuclear activities are not reported, and many stages in the processes are completely
lacking. Until we know these we must hesitate to express an opinion on the position
of Dipodascus.

THE NUCLEUS OF PHYCOMYCETES IN ONTOGENY

A detailed and complete study of the nucleus of some Phycomycetes in the various
phases of ontogeny is greatly to be desired. At present we know a good deal about
the nuclear activities during gametogenesis and something at the time when the
oospore germinates, but the data are not complete for any one form and do not explain
the most important problems of ontogeny. These concern the significance of the
mitoses in the gametangia, the relative numbers of chromosomes at different periods
of ontogeny, and their bearing on the sequence of generations, which is not well
understood in this group.

This knowledge will deraand the study of one or more types with attention to
nuclear phenomena during vegetative periods, especially at the time when asexual
spores or conidia are formed, during gametogenesis and the mitoses following the
fusion of sexual nuclei. Species of Albugo and Peronospora seem to offer the best
material for these investigations. Pythium, although easy to cultivate and control,
has nuclei so small as to be almost impossible for such details, and the same difficul-
ties apply to the Mucorales and in part to the Saprolegniales, while in this latter
group the complications of apogamy render the forms useless for these problems.
Speculations on the reduction of chromosomes and the significance of various phases
of ontogeny in this group are almost futile until we have convincing and complete
data for one or more types.

Whatever may be the significance of the mitoses in the gametangia, there is no
proof that they are reducing divisions, and it is probable that they are only phylo-
genetic reminiscences. Stevens's observations that the nuclei in the second mitosis of
Albugo are much weaker in kinoplasm are interesting, but it is very questionable
whether such divisions are necessary steps in the physiological differentiation of
gametes. The mitosis may have simply a phylogenetic relation and the lessened
kinoplasmic content be merely the result of that decrease in the size of the nuclei
characteristic of advanced periods of oogenesis in these plants.

Everything seems to point to the ooplasm as trophoplasmic in character, first
from the gathering of substance around the ccenocentrum,' and second from the effect

251

28 Oogenesis in Saprolegnia

of this structure on nuclei in the vicinity. Staining reactions confirm this conclusion,
but it is not wise to lay too much stress on the efPects of stains in objects so small as
these. And for this reason the judgment that the gamete nuclei are weak in kino-
plasm must be taken with caution. The nuclei are generally smaller, and the condi-
tions are such that the majority of them must disorganize ; but the reason for this
run-down state must be chiefly the general nutritive conditions of the gametangium,
and not the mitoses of that period.

Stevens (1901, pp. 238, 239) lays stress on that period of oogenesis in Albugo
and Peronospora termed " zonation," when the nuclei often in mitosis move from the
center of the oogonium to the periphery. He suggests "that the nuclei pass to
the periphery to rid themselves of superfluous kinoplasm, possibly to prevent partheno-
genetic development of the oosphere." This theory seems to the writer to suppose
an order of events and degree of preformed specialization more intricate than the
evidence warrants. It seems more likely that "zonation" represents an event that
happens to accompany, but is secondary to, those processes which gather the ooplasm
in the center of the oogonium and give the egg its coenocentrum and characteristic
alveolar structure.

Indeed, the conditions that cause that extraordinary degeneration of nuclei in
the oogonium must furnish in large part the solutions of these problems. This
phenomenon is universal whether the nuclei break down in the eggs themselves
(Saprolegniales) or are relegated to such secondary sexual structures as periplasm
(Peronosporales) or a conjugating tube, as in the Ascomycete Pyronema. As we have
seen in Saprolegnia, the many nuclei in the eggs during advanced stages of oogenesis
are all much reduced in size, and the only thing that saves the fortunate survivors of
the generally severe conditions is proximity to that center of metabolic activity, the
coenocentrum. There is a limit to the number of nuclei possible in a given amount of
cytoplasm. The nutrition of the oogonium decreases as oogenesis proceeds, and finally
reaches a point when the nuclei are sorely pressed to maintain themselves. This cer-
tainly seems to be the history for the Saprolegniales, and probably every Phycomycete
whose sexual organs are coenogametes, as these structures are generally formed late in
ontogeny when the period of vegetation is about completed.

The nuclei are then subjected to a keen struggle for existence, and, in spite of
the fact that they are in a symplast, which is itself a unit, they may well be supposed
each to look after its own interest as far as possible. The outcome of that struggle is
largely determined by the activities of the cytoplasm, which may develop such metabolic
centers (morphologically expressed by ccenocentra) that certain nuclei by good fortune
of favorable position are given great advantages over their neighbors and finally
selected as the survivors.

There are a number of instances known where structures sacrifice some of their
nuclei to provide the remainder with the cytoplasm at hand. Certain of the Fucales
are notable examples, and there will probably be found other illustrations among the

252

Bradley Moore Davis 29

algae and fungi. Analogous conditions in animals have been reported, as in oogenesis
of Actinosphseria (Hertwig, 1898), and the well-known fate of supernumerary male
nuclei in polyspermy. In these cases there has not been reported the same close rela-
tion between the surviving nuclei and metabolic centers of the cell as between the
favored nuclei and the ccenocentra of the Saprolegniales and Peronosporales. In
this same connection we need more detailed accounts of oogenesis in the Fucales,
Vaucheria and Sphseroplea.

The reasons why the oogonium overstocks itself with nuclei are probably phylo-
genetic and recall the time when numerous uninucleate gametes were formed from the
protoplasm that now acts as a unit (coenogamete). Such uninucleate gametes were
probably smaller than their homologues, the asexual zoospores, as is so characteristic
of algse. Among the algsD it is generally conceded that the small gamete swarm
spores result from diif erent conditions of nourishment than their asexual homologues.
It has been suggested that they are starved, but that seems a clumsy conception of
very intricate processes. But there must be deep significance in the overproduction
of sexual nuclei during gametogenesis and its obvious association with the deficient
nutrition at the command of the gametangia. This phase of the subject has not
received the attention it deserves.

SUMMAKY OF THE INVESTIGATION OF SAPROLEGNIA

OOGENESIS

The material, Saprolegnia mixta, was apogamous, being entirely free from
antheridial filaments.

The resting nucleus has a loose linin network and a nucleolus, and presents
essentially the structure of the nucleus of higher plants.

There is one mitosis in the oogonium, the spindle being intranuclear. There are
no centrosomes. The four chromosomes are derived from the linin network.

The daughter-nuclei following the mitosis are much smaller than their parents.
They shortly give evidence of coming degeneration, the nuclear membranes become
indistinct, and the contents finally lie as granular material in clear areas resembling
vacuoles.

The eggs are formed during the process of nuclear degeneration. The protoplasm
in the oogonium at this period is arranged peripherally around a large central vacuole.
The ooplasm collects around several centers, each of which is to become an eg^ origin.
The egg origins are finally separated through the arrangement of vacuoles which
results in the severance of connecting strands of protoplasm, and the eggs round
themselves off as independent structures.

The differentiation of the egg origins takes place around a deeply stained proto-
plasmic body, the ccenocentrum , from which delicate fibrillae radiate. The coenocentrum
is formed de novo, one for each spore origin. It is at first a small globule, made con-
spicuous, however, by its fibrillar rays. It is most conspicuous in the young eggs,

253

30 Oogenesis in Saprolegnia

becoming less distinct with the ripening, and finally disappears. There is no trace of
it in the oldest eggs.

The coenocentrum is a protoplasmic structure, but not a permanent organ of the
cell. It is probably the morphological expression of dynamic activities in the oogonium
when the egg origins are differentiated, and is a sort of focal point of the metabolic
processes peculiar to oogenesis.

The coenocentrum exerts a chemotactic influence on any nuclei in its immediate
vicinity. Generally one nucleus is selected and comes to lie very close to the coeno-
centrum, so that these two structures in the egg origins may be separated only under
high magnification. This nucleus increases in size when all other nuclei in the egg
origins and young eggs are degenerating, showing that it is greatly favored with
respect to nourishment by its position near the coenocentrum.

Sometimes two or even three nuclei may lie sufficiently near the coenocentrum to
be saved from degeneration, and such eggs are in consequence bi- or trinucleate.
Binucleate eggs are not uncommon, trinucleate eggs are more rare.

As the eggs mature, the favored nucleus increases greatly in size, until it is many
times larger than at the period following the mitosis. The other nuclei have gener-
ally completely disorganized, but sometimes traces remain as granules scattered in the
cytoplasm.

Binucleate eggs in the Saprolegniales need have no relation to the problem of
sexuality, and Trow's conclusions are not established.

spobogenesis

A general confirmation of the accounts of Rothert, Hartog, and Humphrey.

The uninucleate spore origins are differentiated by clefts that push their way from
the central vacuole of the sporangium to the periphery. When the clefts reach the
cell wall, the turgor of the sporangium is relieved through the escape of water, and
the spore origins run together, but they soon draw apart and round themselves off as
zoospores. There seem to be no cytoplasmic centers in the sporangium comparable to
the coenocentra.

LITERATURE CITED

Behrens. " Einige Beobachtungen uber die Entwickelung des Oogons und der Oosphare von

Vaucheria." Ber. d. deut. hot. Gesell, Vol. VIII (1890), p. 314.
Davis. " The Fertilization of Albugo Candida." Bot. Gaz., Vol. XXIX (1900), p. 297.
GoLENKiN. " Ueber die Befruchtung bei SiJhaeroplea annulina und tiber die Structure der

Zellkerne bei einigen griinen Algen." Bull. Soc. Imp. Nat. Moscow, 1899, p. 343.
Gbubeb. "Ueber das Verhalten der Zellkerne in den Zygosporen von Sporodinia grandis,

Link." Ber.d. deut. bot. Gesell, Vol. XIX (1901), p. 51.
Harper. " Sexual Reproduction in Pyronema confluens and the Morphology of the Ascocarp."

Ann. of Bot., Vol. XIV (1900), p. 321.

254

Bradley Moore Davis 31

Habtog. "Recent Researches on the Saprolegniaceae : A Critical Abstract of Rothert's

Results." Ibid., Vol. II (1888), p. 201.

"Some Problems of Reproduction." Quar. Jour, of Mic. Set., Vol. XXXIII (1891).

" On the Cytology of the Vegetative and Reproductive Organs of the Saprolegnieae."

Trans. Roy. Irish Acad., Vol. XXX (1895), p. 645.

" The Cytology of Saprolegnia." Ann. of Bot, Vol. X (1896), p. 98.

" The Alleged Fertilization in the Saprolegnieae." Ibid., Vol. XIII (1899), p. 447.
Hektwig. " Ueber Kerntheilung, Richtungskorperbildung und Befruchtung von Actinosphae-

rium eichorni.^' Abhand. bayer. Akad. Wiss., Vol. XIX (1898).
Humphrey. " The Saprolegniaceae of the United States, with Notes on Other Species." Amer.

Phil. Soc, 1892.
JuEL. "Ueber Zellinhalt, Befruchtung und Sporenbildimg bei Dipodascus." Flora, Vol. XCI

(1902), p. 45.
Klebahn. "Studien liber Zygoten; II: Die Befruchtung von Oedogonium Boscii." Jahrb.

f. wiss. Bot., Vol. XXIV (1892), p. 235.

" Die Befruchtung ron Sphaeroplea annulina Ag." Festschrift fiir Schwendener, 1899, ji. 81 .
Klebs. " Zur Physiologic der Fortpflanzung einiger Pilze; II: Saprolegnia mixta." Jahrb. f.

wiss. Bot., Vol. XXXIII (1899), p. 71.
MiTZEEWiTscH. " Ucbcr die Kerntheilimg bei Spirogyra." Flora, Vol. LXXXV (1898), p. 81.
MiYAKE. " The Fertilization of Pythium de Baryanum." Ann. of Bot., Vol. XV (1901), p. 653.
Nichols. " The Morphology and Development of Certain Pyrenomycetous Fungi." Bot. Gaz.,

Vol. XXII (1896), p. 301.
Oltmanns. " Ueber die Entwickelung der Sexualorgane bei Vaucheria." Flora, Yo\. LXXX

(1895), p. 388.
RoTHERT. " Die Entwickelung der Sporangien bei den Saprolegnieen." Cohn's Beitrdge z.

Biol.d. Pflan., Vol. V (1888), p. 291.
Ruhland. " Die Befruchtung von Albugo lepigoni und einigen Peronosporaceen." Hedwigia,

Vol. XLI (1902), p. 179.
Stevens. " The Compound Oosphere of Albugo bliti." Bot. Gaz., Vol. XXVIII (1899), p. 149.

" Die Gametogenese und Befruchtung bei Albugo." Ber. d. deut. bot. Gesell., Vol. XIX

(1901), p. 171.

" Gametogenesis and Fertilization in Albugo." Bot. Gaz., Vol. XXXII (1901), p. 77.

"Studies in the Fertilization of Phycomycetes : Sclerospora." Ibid., Vol. XXXIV (1902),

p. 420.
Strasburger. " Schwarmsporen, Gameten, pflanzlichen Spermatozoiden und das Wesen der

Befruchtung." Hist. Beitrdge, Vol. IV (1892), p. 48.
Thaxter. " New or Peculiar Aquatic Fungi; I: Monoblepharis." Bot. Gaz., Vol. XX (1895), p. 433.

"New or Peculiar Aquatic Fungi; II: Gonopodya Fischer and Myrioblepharis nov. gen."

Ibid., p. 477.

"New or Peculiar Aquatic Fungi; III: Blastocladia." Ibid., Vol. XXI (1896), p. 45.

"New or Peculiar Aquatic Fungi; IV: Rhipidium, Sapromyces, and Araiospora." Ibid.,

p. 317.
Timberlake. "Development and Structure of the Swarmspores of Hydrodictyon." Trans.

Wis. Acad. Sci. Arts and Letters, Vol. XIII (1902), p. 486.
Tbow. " The Karyology of Saprolegnia." Ann. of Bot, Vol. IX (1895), p. 609.

" Observatfons on the Biology and Cytology of a New Variety of Achlya Americana."

Ibid., Vol. XIII (1899), p. 131.

" Biology and Cytology of Pythium ultimum." Ibid., Vol. XV (1901), p. 269.

255

32 Oogenesis in Sapeolegnia

Wager. " Observations on the Structure of the Nuclei of Peronospor a parasitica" Ibid., Vol.

IV (1889), p. 127.

" On the Structure and Reproduction of Cystopus candidus Lev." Ibid., Vol. X (1896),

p. 295.

"On the Fertilization of Peronospora parasitica." Ibid., Vol. XIV (1901), p. 263.
WissELiNQH. " Ueber Kerntheilung bei Spirogyra." Flora, Vol. LXXXVII (1900), p. 355.

EXPLANATION OF PLATES

The material, fixed in weak chrom-acetic acid, was cut 3 /x thick and stained on the slide
with safranin and gentian violet. All figtires were sketched with an Abbe camera imder the
Zeiss apochromatic objective 2 mm. aper. 1.30, or 1.5 mm. in combination with compensating
oculars. The magnification is as follows: Fig. 1, 250 diameters; Figs. 2-5, 12, and 13, 500
diameters; Figs. 6-9, 1,500 diameters; Figs. 10, 11, and 14-29, 1,000 diameters; Figs. 30-35, 667
diameters.

PLATE XV
(Figs. 1-15 illustrate Oogenesis)

Fig. 1. — End of hypha about to form an oogonium.

Fio. 2. — Young oogonium, nuclei approaching spirem.

Fig. 3. — Central vacuole forming.

Fig. 4. — Central vacuole; nuclei in spirem.

Fig. 5. — More advanced than Fig. 4; nuclei in mitosis.

Fig. 6. — Details of nucleus in spirem condition.

Fig. 7. — Metaphase of mitosis; spindle intranuclear; nucleolus outside of spindle; three
chromosomes shown.

Fig. 8. — Mitosis just after the splitting of chromosomes at nuclear plate; nucleolus outside
the spindle.

Fig. 9. — Anaphase; two groups of chromosomes, four in each group, at the poles.

Fig. 10, — Oogonium after mitosis with twice as many nuclei as previous to that event.

Fig. 11. — Oogonium older than in Fig. 10; nuclei degenerating. *

Fig. 12. — Formation of egg origins under low magnification (500 diameters); ccenocentrum
in center of each egg origin.

Fig. 13. — Egg origins older than in Fig. 12; coenocentra with conspicuous radiations.

Fig. 14. — Ccenocentrum before the differentiation of the egg origins; radiations plain;
nucleus at one side of ccenocentrum; other nuclei degenerating in the cytoplasm.

Fig, 15. — Similar to Fig, 14; two coenocentra; nucleus at the side of each; many nuclei
degenerating in the cytoplasm,

PLATE XVI
(Figs. 16-29 illustrate Oogenesis)
Fig. 16. — Egg origin just before rounding off to form egg; conspicuous ccenocentrum with
nucleus at the side.

Fig. 17. — Young egg, nucleus larger than in Fig. 16.

Fig. 18.^Young egg; ccenocentrum without radiations.

Fig. 19. — Egg, older than in Figs. 17 and 18; nucleus larger.

Fig. 20.— Egg with nucleus extended toward ccenocentrum, which has almost disappeared.

Fig. 21. — Mature egg; large nucleus; ccenocentrum disappeared.

Fig. 22. — Young binucleate egg, the two small nuclei close to the smaller ccenocentrum.

256

Bradley Moore Davis 33

Fia. 23. — Young egg with two coenocentra, each accompanied by a nucleus.
Fig. 24. — An exceptionally large binucleate. egg with prominent coenocentrum.
Fig. 25. — Egg with two nuclei lying over one another, both extended toward the coeno-
centrum.

Fig. 26. — Binucleate egg with the nuclei at a distance from one another.

Fig. 27. — Binucleate egg with the nuclei close together.

Fig. 28. — Trinucleate egg, the three nuclei lying close together.

Fig. 29. — Trinucleate egg, the three nuclei at a distance from one another.

(Figs. 30-35 illustrate Sporogenesis)
Fig. 30. — End of sporangium showing development of central vacuole.
Fig. 31. — Portion of cross section of sporangium, central vacuole well developed.
Fig. 32. — Early stage of segmentation; cleavage furrows running from central vacuole to
periphery.

Fig. 33. — After cleavage furrows have reached periphery, spore origins forming.
Fig. 34. — Spore origins older than in Fig. 33.
Fig. 35. — Zoospores in sporangium.

257

.**

Decennial Publications, X

Plate XV

iavis.iel

Lith Anst.-v;E.AJu-n'ke,Leipziq

Decennial Publications, X

Plate XVI

Davis del.

Lith.AiistArE.A.Ftmlte,leipzicr.

^ri R A R r

C.ALIFO^^

14 DAY USE

RETURN TO DESK FROM WHICH BORROWED

LOAN DEPT.

This book is due on the last date stamped below, or

on the date to which renewed.

Renewed books are subject to immediate recall.

jUN2ri9668^

Ml^^^

RCO

APR 30 1968

RECEI VE D

MAY T'On-iOA i Vl

LOAT-^ D^PT.

MAY 211969 9 5

l. i 'I'T' TO TKh'ilLi

SCIENCES LIB.

JUN 3 1969
REC'P L.D

R 51974 8 9

T T) 21A-60w-10,'65
777638l0)476B

General Library

University of California

Berkeley

'fimafl^Bxi

■•'V^.i*'