The ultrastructure and phylogenetic significance of avian spermatozoa

The Ultrastructure and Phylogenetic Significance
of Avian Spermatozoa

By

ROBERT W. MciFARIANE

A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR, THE D:EGREP OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA
1971

ACKNOWLEDGMENTS

This research was supported by grants-in-aid from

the Frank M. Chapman Memorial Fund of the American Museum

of Natural History, the Society of the Sigma Xi and RESA

Research Fund, and the Organization for Tropical Studies.

I am grateful for the support provided by the Bowdoin

Scientific Station, Kent Island, New Brunswick; the Carib-

bean Conservation Corporation, Tortuguero, Costa Rica; and

the Pacific Ocean Biological Survey Program, Smithsonian

Institution.

The many individuals who have assisted in the field

work of this study are unfortunately too numerous for

individual recognition. I can only hope that their memories

of our offtimes brief association are as pleasant as mine.

I wish to extend my deepest appreciation to Professor

Pierce Brodkorb for his advice, criticism, and constant

encouragement. I wish to thank Drs. Ernst Kallenbach, Frank

Nordlie, James Gregg, and Carl Feldherr for their criticism

of the manuscript.

Special thanks are due Dr. William P. Callahan, formerly

of the Department of Anatomical Sciences of the University

ii

of Florida. His generous and continuous support in

innumerable ways was totally responsible for the completion

of this research.

iii

TABLE OF CONTENTS

ACKNOWLEDGMENTS . .

LIST OF FIGURES .

ABSTRACT . .

INTRODUCTION . .

METHODS . .

ULTRASTRUCTURE . .

The Head . .

Nucleus . .

Acrosome . .

Helical Membrane .

Apical Spine or Body

The Tail .

Axoneme . .

Neck . .

Middle Piece .

. o *

. o o S .

. o .

. o .

. 5 0

. 5 .

. . .

. . .

. . .

. o o

. . o

. o

. . .

* S i

. o .

Principal Piece and End Piece

Cytoplasmic Membrane and Annulus

. . 23

SI

S1

S

S

S

S

S

Si

S

S

S

S

Page

ii

vi

viii

1

3

5

6

7

10

11

13

14

14

16

19

22

O

A COMPARISON OF AMNIOTE SPERMATOZOA .

The Head . . . .

The Tail . . . .

A COMPARISON OF AVIAN SPERMATOZOA .

SUMMARY . . . .

APPENDIX A. FIGURES 1 73. . .

Key to Abbreviations . .

APPENDIX B. CLASSIFICATION AND LIST OF
STUDIED . . .

LITERATURE CITED . . .

BIOGRAPHICAL SKETCH . . .

Page

24

25

26

31

. 0 0 .
* 0 0

o 0 0 5 5 5

SPECIES

* 0 5 0 6 5
. . .

o .
. .

LIST OF FIGURES

Figure

1.

2-4.

5.

6-11.

12.

13-15.

16-21.

22.

23-24.

25-31.

32.

33-36.

37-40.

41.

42-45.

46-50.

51.

The spermatozoon of Alectoris graeca .

Alectoris graeca .. . . .

The spermatozoon of Sterna fuscata . .

Sterna fuscata . . .

The spermatozoon of Columba livia .

Columba livia . . .

Columba livia . . .

The spermatozoon of Centurus carolinus .

Centurus carolinus . . .

Centurus carolinus . . .

The spermatozoon of Myiarchus crinitus .

Myiarchus crinitus . . .

Tyrannus verticalis . .

The spermatozoon of Tachycineta thalassina.

Tachycineta thalassina . .

Tachycineta thalassina . . .

The spermatozoa of Parus bicolor and

Vireo olivaceus . . .

52-55. Parus bicolor and Vireo olivaceus . .

vi

Page

42

44

46

48

50

52

54

56

58

60

62

64

66

68

70

72

74

76

Figure

56. The spermatozoa of Turdus migratorius

and Piranga rubra . . .

57-59. Turdus migratorius, Piranga rubra, and

Tangara gyrola . . .

60. Testicular section of Piranga rubra .

61-63. Testicular sections of Piranga rubra .

64-65. Testicular sections of Pipilo ery-

throphthalmus . . .*

66-73. Some representative sperm . . .

vii

Page

78

80

82

84

86

88

Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy

THE ULTRASTRUCTURE AND PHYLOGENETIC SIGNIFICANCE
OF AVIAN SPERMATOZOA

By

Robert W. McFarlane

August, 1971

Chairman: Dr. Pierce Brodkorb
Major Department: Zoology

Avian spermatozoa exhibit considerable variation in

both gross and fine structure and may be useful indicators

of phylogenetic relationship but knowledge of their ultra-

structure is required. The spermatozoa of 281 species of

birds were surveyed with phase-contrast microscopy. Of

these, 177 species were viewed with the electron microscope

and the ultrastructure of 18 species was studied in detail.

The head is typically cylindrical or fusiform, and

the nucleus contains granular chromatin limited by a double

membrane. In non-passerines the acrosome is conical with a

blunt or pointed tip and limited by a single membrane. In

passerines it has a lateral extension forming a helix and is

limited by a double membrane. Subacrosomal structures include

apical spines, spheres, and granular material in some, but

not all, species.

The tail has a typical 9+2 flagellum. Accessory fibers

are known for only two orders. The fibers are very small in
viii

Galliformes but well developed in Passeriformes. They are

cylindrical, uniform in size, remain attached to the peri-

pheral doublet tubules at maturation, and terminate at a

common point on the distal axoneme. The proximal centriole

of the neck is adjacent to the nucleus, at approximately a

right angle to the flagellar axis. The implantation fossa

may be shallow, deep, or tangential. The distal centriole

is continuous with the peripheral doublets of the axoneme,

whose central tubules arise at its distal margin. Pas-

serines apparently lack the proximal centriole. Instead,

the anterior ends of the accessory fibers surround the dis-

tal centriole and form a connecting piece.

The mitochondria of the middle piece may be oblate or

polygonal, loosely grouped or highly organized, and vary in

number from one to more than 1000. The mitochondrial cristae

are parallel to and thicker than the outer mitochondrial

membrane. Intermitochondrial granules are found in Columbi-

formes. The middle piece of non-passerines composes 5 to

75 per cent of the tail. Passerines have a single helical

mitochondrion, which may extend along the tail for 10 to 90

per cent of its length. Some passerines have a granular

substance of unknown origin or function around the proximal

axoneme.

It is difficult, and perhaps unnecessary, to distinguish

between the principal piece and the end piece of the tail.

ix

The highly structured fibrous sheath of reptiles and mammals

is absent in avian sperm. The Galliformes and Tinamiformes

have an amorphous sheath, but in all other species the

axoneme is encased in only the cytoplasmic membrane. An

annulus is present in Gallus but has not been observed in

other species.

The spermatozoa of five avian orders whose ultrastruc-

ture is known can be differentiated on their fine structure.

The most distinctive is the helical configuration of the

Passeriformes, which demonstrates a complete evolution from

a small, slightly coiled form in primitive families to a

large, elongate, and markedly helical shape in those families

most recently evolved. The helical configuration, whose

functional significance remains unknown, has independently

evolved in two other orders. The helical spermatozoa of the

Scolopacidae are quite distinct from those of six other chara-

driiform families. A helical acrosome occurs in Procel-

lariidae but is lacking in two other procellariiform families,

the Dio'.edeidae and Hydrobatidae.

The spermatozoon of the Columbiofrmes is also distinct

from any other of the twenty orders which have been examined,

principally because of its middle piece. The Piciformes

have a generalized sperm which superficially resembles that

of several other orders.

INTRODUCTION

In recent years numerous investigations have revealed

the fine morphological details of representative spermatozoa

for most animal phyla, and a number of phylogenetic compari-

sons have been made (Baccetti, 1970). My objectives were to

determine the extent of morphological variation among the

spermatozoa of the Class Aves and to identify any morpholo-

gical characteristics which might serve as indicators of

phylogenetic relationship among the various orders and

families of birds. An earlier study (McFarlane, 1963) had

demonstrated the feasibility of such comparison.

Sperm samples of 281 species of birds were collected

and surveyed with phase-contrast microscopy (appendix B).

All spermatozoa exhibit variations upon a basic cylindrical

form, generally of 0.5p diameter or less. Inasmuch as the

limit of resolution with light optics is 0.2p, much of the

fine detail is lost or obscured by diffraction images. The

greater resolution and depth of field of electron microscopy

were utilized to study the gross morphology of 177 species.

Most knowledge of the ultrastructure of vertebrate

spermatozoa is based upon numerous investigations of mammalian

1

2

species, while birds have been virtually ignored. Only one

avian species, the domestic fowl, Gallus gaj.lus, has been

adequately studied. Nagano (1962) described the fine

structure of the developing spermatid, Nicander and Hellstrom

(1967) studied mitochondrial changes, McIntosh and Porter

(1967) described the role of microtubules in effecting

changes in the shape of the head during spermiogenesis, and

Lake et al. (1968) described the ultrastructure of the mature

fowl spermatozoon. Studies of the fine structure of the

tail for the house sparrow, Passer domesticus,were contri-

buted by Sotelo and Trujillo-Cenoz (1958), and fcr the tree

sparrow, Passer montanus,and the munia, Uroloncha (=Lonchura)

striata, by Yasuzumi (1956a) and by Masuda (1958). Yasuzumi

(1956b) also commented on the structure of the nucleus in the

spermatid of the tree sparrow, Passer montanus, Furieri

(1962-63) described the mature spermatozoa of the chaffinch,

Fringilla coelebs, and Nicander (1968, 1970) illustrated

some aspects of the spermatozoon of the zebra finch,

Taenijo.gia castanotis (=Poepila guttata).

Thus all references to the ultrastructure of avian

spermatozoa are based upon only six species, limited to

only three of the 177 living families of birds (Wetmore,

1960). I have investigated the ultrastructure of an

additional 18 species of birds (appendix B).

METHODS

The preferred fixatives for ultrastructural investi-

gations are osmium tetroxide and glutaraldehyde. Both of

these chemicals deteriorate rapidly at normal environmental

temperatures, and osmium requires that dehydration quickly

follow fixation. Optimum preservation of fine structure

requires that fixation occur at near-freezing temperature.

The collection of spermatozoa from an adequately large

number of species for comparative study necessitates

fixation under field conditions where refrigeration is

unavailable or impractical. My comparative tests demon-

strated that phosphate-buffered formaldehyde was stable at

all temperatures, and fresh tissue preserved in this fixative

could remain immersed for extended periods without deterio-

ration or further processing, even for several years. While

the quality of such fixation does not equal that of osmium

treatment, it is a satisfactory compromise and was used for

the majority of the specimens collected.

In a few instances, wild birds were captured and taken

alive to the laboratory, or tissues were fixed in the field

in 1 per cent phosphate-buffered osmium in an ice bath and

3

.4

quickly returned to the laboratory for further processing.

Other specimens initially fixed in 10 per cent phosphate-

buffered formalin were subsequently post-fixed in osmium to

utilize the staining properties of the metal. All tissues

used for ultrastructural studies were dehydrated in a

graduated ethanol and propylene oxide series and embedded

in Araldite epoxy. Sections were cut with glass knives and

stained with lead citrate, uranyl acetate, or vanadium

molybdate.

Unsectioned sperm were supported by a parlodion film

over a copper grid. Some specimens were shadowed with

chromium to effect a third dimension in the electron

micrographs. All specimens were observed in a RCA EMU-3C

electron microscope at 50 kV and recorded on sheet film.

The spermatid undergoes the final stages of maturation

while traversing the vas deferens, a reduction in head size

and increase in tail length. Therefore, all descriptions

of mature spermatozoa are based upon sperm taken from the

distal portion of the vas deferens, or the seminal glomerule.

In a few instances, sections of testicular material were made

(see fig. 60-63) to aid in identification of the various

spermatozoan components. No attempt was made to determine

the details of spermiogenesis for any species.

ULTRASTRUCTURE

The function of any spermatozoon is the transmission

of the hereditary material, the chromatin, to a complemen-

tary egg cell, and gaining entry into the egg once encoun-

tered. Thus a sperm has two very different problems to

solve: it requires a means of locomotion to propel it to-

wards the egg, and a mechanism to penetrate the several mem-

branes surrounding the egg in order that the male chromatin

may enter into the cytoplasm of the egg. The propulsive

force is provided by the primary component of the tail, the

axoneme and its associated structures, which include a

source of energy. Egg penetrance is the task of the head,

which contains the chromatin within its nucleus, and is

achieved through the little understood role of the acrosome

and its associated structures.

Most knowledge of the ultrastructure of vertebrate

spermatozoa is based upon mammalian species, and some degree

of standardization of the nomenclature has resulted (Fawcett,

1965; Hancock, 1966). This terminology is generally appli-

cable to birds, but certain ambiguities result, particularly

concerning the term "head". The head of a spermatozoon con-

5

6

sists of two major structures, the nucleus and the acrosome.

The tail of a sperm has four distinct regions: a very small

neck, attaching the tail to the nucleus; the middle piece,

where a sheath of mitochondria encircles the axial filament;

the principal piece, with the axial filament surrounded by

a protein sheath; and the end piece, where the axial fila-

ment is covered only by the cytoplasmic membrane.

The juncture of the head and tail of mammalian sperm

is easily determined for the heads tend to be large and

spatulate, and the tails slender and cylindrical. However,

the heads of all avian sperm are basically cylindrical or

fusiform, so that it is usually difficult to determine the

juncture between the nucleus and the middle piece. Fre-

quently, the most conspicuous juncture is between the middle

piece and the principal piece. Thus what first appears to

be the head actually includes the anterior segment of the

tail. Head and tail are still very useful terms and in

view of their long acceptance it seems unnecessary to employ

new terms. But the potential ambiguity must be remembered

in comparisons with non-avian species.

The Head

The structures which may be observed as components of

the head include the nucleus, the acrosome, the apical

spine or apical body, and the helical membrane which is

7

associated with the acrosome of many species.

The Nucleus

The nucleus, which contains the chromatin of the gamete,

is the raison d'etre for all other sperm components. It is

derived from the nucleus of the spermatid. During spermio-

genesis the spermatid nucleus undergoes a change in shape

and texture from a round, homogeneous body to a smaller

dense body with short strands of tightly packed chromatin

(fig. 60), and finally achieves a very dense and granular

appearance. In the mature spermatozoon the nuclear material

is rarely homogeneous and frequently contains a number of

small irregular cavities randomly distributed (fig. 52).

These cavities are not membrane-limited and cannot be consi-

dered true vacuoles. They are found in the spermatozoan

nuclei of many non-avian species and are believed to result

from incomplete condensation of the chromatin.

The nucleus is typically limited by a double membrane

(fig. 7, 16, 26, 31). Excess nuclear membrane accumulates

in the neck region of Gallus (Lake et al., 1968), similar

to the scrolls described in mammals (Fawcett, 1970). This

has not been observed in other avian species.

The chromatin of the mature spermatozoon is electron

opaque and very resistant to impregnation by the embedding

medium. Thin sections of embedded material occasionally

8

split under the electron beam, and this split invariably

begins in the nuclear area (note the small holes in the

nucleus of fig. 18). The nucleus of Columba exhibits

large, electron-dense strands in a semi-transparent matrix

(fig. 13, 16). This may be an artifact but it has also been

observed by light microscopists (Smith, 1912; Mehrota, 1951)

using several different fixatives.

The nucleus of most non-passerines is a straight

cylinder. That of Gallus has a definite curvature, not

necessarily planar (McIntosh and Porter, 1967). Passerine

nuclei tend to reflect the helical nature of the acrosome

and middle piece. The wave length of the helix is highly

variable and characteristic of the family. The nuclei of

some species deviate only slightly from a straight axis.

In others the nucleus may complete one or more helical

revolutions. Some species have a helical constriction which

resembles a furrow, spiralling the length of the nucleus

(fig. 52). In general, the nucleus is the most conservative

of all sperm components, exhibiting the least variation when

inter-specific comparisons are made (Allen et al., 1968).

McIntosh and Porter (1967) have described two sequential

sets of microtubules that effect the final nuclear shape in

Gallus. The first set is a thin helical band, only one

tubule deep, which effects the shrinkage and transformation

9

of the spherical nucleus into its elongate form. This set

disappears and is replaced by the manchette, straight

tubules lying parallel to the nucleus and extending from

the acrosome to the anterior tail. The manchette is also

lost during maturation.

Nicander (1967, 1970) states that only one transitory

set of microtubules is found in Taeniopygia (=Poephila).

This occurs as a twisted bundle which includes cisternae

of endoplasmic reticulum. It winds around the nucleus and

proximal tail during formation of the helix and is shed

during final maturation. Sotelo and Trujillo-Cenoz (1958)

describe similar tubules for Passer.

The fixatives utilized in this study are notoriously

deficient for the preservation of microtubules, which are

best preserved with glutaraldehyde (McIntosh and Porter,

1967). Nevertheless, microtubules have been observed in

Passerina, Pipilo, and Tachycineta after Os04 fixation.

They occur as a spindle-shaped, membrane-limited bundle

extending from the acrosome to the tail and are seen only

in the developing spermatid (fig. 60-65). During their

formation they incorporate cisternae of the endoplasmic

reticulum (fig. 64). The cisternae subsequently collapse

and remain as isolated membranes within the bundle (fig.

62-65). The bundles are apparently sloughed off during final

10

maturation and are conspicuous in the cellular debris found

in the vas deferens.

In one favorable section of the spermatid of Thryothorus

a small number of microtubules can be seen adjacent to the

acrosome. This raises the possibility that proper glutaral-

dehyde fixation might reveal other microtubules similar to

those observed in the fowl. The asymmetrical microtubule

spindle, which is believed to effect the helical configuration

of the spermatozoon, would not appear to be sufficient to

effect the initial shrinkage and elongation of the nucleus.

The Acrosome

Nagano (1962) and McIntosh and Porter (1967) have

described the development of the acrosome from the Golgi

complex in Gallus. It attaches to the nucleus as a spheri-

cal body, then elongates to its final form. The acrosome

is the most variable of all sperm components, both in size

and shape. In non-passerines it remains much smaller than

the nucleus and is essentially conical, with a pointed or

blunt anterior tip. Crypturellus has a spade-like anterior

projection (fig. 68). The plane of contact between the

acrosome and the nucleus may be perpendicular or oblique

to the longitudinal axis. Among the passerines, the size

varies from a small minor component to a very large struc-

ture, which dominates the head and is several times larger

than the nucleus.

11

The acrosome is limited by what generally appears to

be a single membrane. However, in favorable sections the

acrosomes of three passerines (Myiarchus, Pipilo, and

Piranga) are definitely limited by a double membrane (fig.

34, 60, 61).

The acrosomal material of non-passerines appears to be

uniform throughout, although it frequently has a granular

composition (fig. 8). Membrane-limited vacuoles are present

in the acrosome of Centurus (fig. 25, 27), but have not been

observed in other species. Among those passerines with an

extensive helical membrane, the acrosome appears to have a

dense core, which is not membrane-limited, and has a less

dense matrix at the base of the helix (fig. 42, 45).

The Helical Membrane

This structure is a lateral extension of the acrosome

forming a left-handed helix that is highly variable in both

length and width. It is a feature of all passerine species

and has arisen independently in at least one family of two

other orders (Procellariiformes and Charadriiformes). In some

instances it is restricted to the very tip of the acrosome

but in most of the passerines it extends the full length.

In Myiarchus the acrosome is relatively small and of

uniform density throughout (fig. 34). In Tachycineta, the

acrosome is large, being three times longer than the nucleus

12

and a sagittal section (fig. 42) reveals a dense core, a

less dense matrix, and dense material at the lateral extre-

mity of the membrane. In transverse section (fig. 45) the

dense extremity appears less conspicuous and apparently

results from the angle of sectioning. The sagittal section

of Parus (fig. 52) would seem to verify this explanation.

If one examines the developing spermatid of Piranga (fig.

60) it seems obvious that two types of acrosomal material

are present: a dense core, not membrane-limited, and a less

dense asymmetrical matrix. Transverse sections of Piranga

spermatids (fig. 61) demonstrate two conditions. Some of

the acrosomes have a dense core surrounded by a less dense

matrix. An area of intermediate density is found at the

apex of the membrane. Other acrosomes have a homogeneous

density. Longitudinal sections of mature Piranga sperma-

tozoa demonstrate the same discrepancy. Some acrosomes are

completely homogeneous from tip to nucleus, and from the

center to the lateral edge of the membrane, while others

have three distinct zones of different densities. All sec-

tions underwent identical fixation and staining. Apparently

the zonation of early development disappears at full matu-

ration.

Micrographs of whole mounts of the sperm of many

passerine species demonstrate a decreasing density gradient

13

from the core to the lateral edge corresponding to the

decreasing thickness of the membrane, with no indication of

high density material at the lateral extremity (fig. 58).

The Apical Spine or Body

-Some species possess an accessory structure located

between the acrosome and the nucleus. This is known to

occur-in at least three forms: (1) a long spine embedded

within the anterior portion of the nucleus and extending

forward into a cavity of the acrosome, but external to the

limi ing membranes of both structures (fig. 3); (2) a long

spine extending forward from the nucleus into a cavity of

the acrosome but enclosed within the nuclear membrane (fig.

18); and (3) a spherical body partially embedded in the

anterior end of the nucleus and partially enveloped by the

acrosome but external to the limiting membranes of both

(fig. 7-9).

Nagano (1962) has described the formation of the apical

spine-of Gallus from a dense granule which appears in a

cytoplasmic cavity formed at the surface of the nucleus

where the acrosome has attached. Lake et al. (1968) des-

cribe the spine as densely-packed, longitudinally arranged,

layered material, slightly more electron-dense than the

acrosome. They also demonstrate a granular material between

the anterior portion of the spine and the acrosome.

14

The apical spine of Columba (fig. 18) is a projection

of the nucleus, definitely enclosed by the double nuclear

membrane. It is surrounded and separated from the acrosome

by amorphous electron translucent material. The apical body

of Sterna (fig. 7-9) appears to have intimate contact with

the acrosome, and no intervening substance has been observed.

It does not appear to be an integral part of the nucleus.

A similar structure may be present in Sula.

The Tail

Sperm tails are comprised of an axoneme or axial

filament complex and its associated structures. The tail

is normally divided into four distinct areas: the neck,

where the axoneme is attached to the nucleus by a connecting

piece; the middle piece, where the axoneme is surrounded by

mitochondria; the principal or main piece, enclosed by a

protein sheath of some nature; and the end piece, where the

axoneme is enclosed by only the cytoplasmic membrane. These

definitions are based primarily upon mammalian sperm, and

certain difficulties arise in trying to apply them to avian

sperm.

The Axoneme

In non-passerines, the axoneme has the standard pattern

of most flagella, namely a pair of single tubules surrounded

by nine doublet tubules, the so-called 9+2 configuration.

15

The axoneme is derived from the distal centriole of the

spermatid, and the nine peripheral doublet tubules are con-

tinuous with the nine triplet tubules of this centriole. The

two central tubules appear to originate at the posterior end

of the distal centriole (fig. 17).

The unit tubules of the doublets appear differently

in some transverse sections. One of the pair appears hollow,

the other appears as a dense rod with two short arms or

appendages attached (fig. 20). In other sections, parti-

cularly in the posterior axoneme but occasionally in the

middle piece (fig. 49), both members of the pair appear

hollow. In favorable sections a series of lines may be seen

radiating from the center of the complex to each of the

doublets (fig. 21).

In passerines, the axoneme has an additional set of

nine large accessory or dense outer fibers lying distal to

the peripheral tubules, a feature known in a number of non-

avian species and designated the 9+9+2 configuration. These

outer fibers arise as lateral outgrowths of the peripheral

doublets (Fawcett and Phillips, 1970) and secondarily fuse

with a connecting piece at the base of the nucleus (fig. 42,

46-50). In transverse section they are usually circular in

outline and uniform in size. In the proximal axoneme of

some species they are wedge-shaped (fig. 42, 49). They

16

extend from the neck to the principal piece and remain in

direct contact with the peripheral doublets at all times.

In the region of the end piece only 9+2 tubules are present.

Nagano (1962) describes small dense fibers proximally in

Gallus doublets,-but they have not been observed in other

non-passerine species.

The diameter of the axoneme in all species diminishes

in size fromanterior to posterior.

The Neck-

The fine structure of the neck is difficult to ascertain

because of its complexity and the lack of agreement among

authors as to the boundary between the neck and the middle

piece. Nagano (1962) considers both the proximal and

distal centrioles to be included in the neck of Gallus.

Lake et al. (1968) include only the proximal centriole. The

matter is further complicated by comparison with other

vertebrates. ..In mammals, the distal centriole is not present

in mature spermatozoa (Fawcett and Phillips, 1969), and in

snakes a unique structure, the neck cylinder, extends beyond

the posterior end of the distal centriole (Hamilton and

Fawcett, 1968).. For purposes of description, I will consi-

der the avian neck to include both proximal and distal

centrioles, for the axoneme is not complete (it lacks the

central tubules) anterior to the posterior margin of the

distal centriole.

Birds can once again be divided into two groups, the

passerines and the non-passerines, based upon the character-

istics of the neck. In the passerines the dense outer fibers

continue anteriorly to a small connecting piece at the base

of the nucleus and dominate the morphology of the area.

Among the non-passerines studied, those sections in which

the details of the region are not apparent have been labeled

as the centriolar complex.

Nagano (1962) and Lake et al. (1968) have shown that

in Gallus the proximal centriole is located at right angles

to the sperm axis at the base of the nucleus within a

small nuclear cavity, the implantation fossa. The distal

centriole is adjacent to, and perhaps fused with, the

proximal centriole.

In Columba the implantation fossa is but a slight con-

cavity in the base of the nucleus. The proximal centriole

is oriented at an angle of 750 to the sperm axis and fused

to the larger distal centriole (fig. 16, 17, 19).

In passerines, the components of the neck are even less

well known. The only description of ultrastructure during

spermiogenesis for a passerine is that of Sotelo and

Trujillo-Cenoz (1958) for Passer domesticus, and they were

unable to describe the complete sequence of events. They do

18

describe the distal centriole as having direct contact with

the nucleus, and the disappearance of the proximal centriole

at the same time that a juxtanuclear body appears.

The neck of mammalian sperm contains a cross-banded

connecting piece linking the dense outer fibers to the base

of the nucleus (Fawcett, 1970). This structure is absent

in non-passerines, which do not have the outer fibers. It

is present in the developing spermatid of Piranga (fig. 60)

but has not been observed in the mature sperm. Nicander

(1970) indicates that only one modified centriole is present

in mature Taeniopygia (=Poephila) sperm, which apparently

do have a striated connecting piece.

In comparison with non-passerines, the neck of passerine

sperm is very short, may incorporate only one centriole, and

the central tubules of the axoneme extend further anteriorly

(fig. 46, 48). The nine dense outer fibers continue

forward to the base of the nucleus where they terminate in

either a striated or a knob-like body of unknown fine structure,

the connecting piece (fig. 50). The base of the nucleus may

have a shallow implantation fossa (fig. 35, 38, 46), a

deeper concavity (fig. 52), or may articulate tangentially

(fig. 55).

The Middle Piece

This portion of the tail is generally defined as the

segment in which mitochondria surround the axoneme, although

in some species small mitochondria extend into the neck

region.

The middle piece of one of the most variable of sperm

components. The mitochondria may be oblate or polygonal,

loosely grouped or highly organized, and vary in number

from one to more than 1000. The portion of the axoneme

encircled by the mitochondrial sheath varies from less than

5 per cent in Alectoris and Sterna to 75 per cent in Columba.

Nicander and Hellstrom (1967) have demonstrated that

during spermiogenesis in Gallus the inner mitochondrial mem-

branes thicken considerably and eventually become 50 per cent

thicker than the outer mitochondrial membrane. This pheno-

menon has been confirmed by this study (fig. 15) but may

not be universal among birds (fig. 23, 24). The cristae

of the mitochondria are parallel to the long axis, not

perpendicular as in somatic cells (fig. 15).

The mitochondria of non-passerine sperm are much

smaller than those of the somatic cells of the vas deferens

or even of the cellular debris of spermatids and Sertoli cells

found in the lumen of the tubule. The mitochondria of Sterna

retain their oblate shape, and 30 to 40 are loosely arranged

20

in the middle piece (fig. 6, 11). Lake et al. (1968)

describe the mitochondria of Gallus as rectangular plates

curved along the longitudinal axis, with approximately 30

arranged helically around the axoneme in a single helix.

In Centurus approximately 70 polygonal mitochondria are

tightly packed in a single helix about 600 from the longi-

tudinal axis (fig. 23, 24), with five mitochondria required

to complete a revolution around the axoneme (fig. 27).

The middle piece in Columbiformes is the most extreme

of all birds studied. These sperm are the longest of all

non-passerine species, and the middle piece encircles 75

per cent of the axoneme. The mitochondria are polygonal

plates, with five rows forming a shallow 30 helix (fig.

13, 14) around the axoneme. Approximately 1300 mitochondria

are present in the middle piece of Columba.

Columba has an additional structure, not observed in

any other avian species, in the form of numerous small

dense granules interspersed between the mitochondria (fig.

13, 15, 21). These granules may be analogous to the pleo-

morphic dense plaques described in snakes (Hamilton &

Fawcett, 1968).

The middle piece of passerines is unique in that a

single, greatly elongated mitochondrion extends as an

asymmetrical left-handed helix. It may encompass almost

the entire length of the tail (fig. 57, 59) or be relatively

21

short and complete only a single revolution.

In Myiarchus and Tyrannus the mitochondrion is kidney-

shaped in transverse section (fig. 38-40). In Tachycineta

the helical mitochondrion is triangular in transverse section

proximally, cylindrical distally, and has a lateral exten-

sion which encircles the axoneme for 2700 at the base of

the neck (fig. 48). One or two spherical mitochondria are

also present at the base of the neck. The Turdidae have

a very large helical mitochondrion which results in a helix

of greater diameter than the helical membrane of the acro-

some (fig. 57).

The derivation of helical mitochondria is poorly under-

stood. In the spermatid, mitochondria are already consider-

ably larger than the mitochondria of adjacent Sertoli cells

(fig. 60). There is some indication that the large size

results from the coalescence of smaller mitochondria (Masuda,

1958; Mehrota, 1951).

The anterior portion of the middle piece of some pas-

serines contains a unique granular substance surrounding the

axoneme (fig. 42, 46, 47). In some species (Tachycineta,

Parus, Vireo) the mitochondrion extends forward to the neck,

and the granular substance occupies the area around the

mitochondrion and the axoneme for a short distance caudad.

In other species (Thryothorus, Parula, Piranga, Pipilo,

22

Junco, Richmondena, Passerina) the granular substance com-

pletely occupies the middle piece adjacent to the neck, and

the mitochondrion begins several microns distal to the neck.

The granular substance is conspicuous in the spermatid (fig.

60) but remains of unknown origin and function.

The Principal Piece and the End Piece

Among mammals and reptiles, that section of the tail

caudad to the middle piece is encased in a fibrous sheath,

generally consisting of a series of small, dense transverse

ribs attached to two longitudinal columns. This sheath ex-

tends over most of the flagellum and distinguishes the

principal piece of the tail from the end piece, where the

axoneme is encased in only the cytoplasmic membrane.

A highly structured sheath of this nature is absent

in all birds. In only two orders, Galliformes and tinami-

formes, has a sheath of any nature been found. An amorphous

sheath of medium density as seen in Gallus (Nagano, 1962;

Lake et al., 1968) is also present in Alectoris (fig. 2).

A similar sheath can be seen in the tinamous Crypturellus

and Nothoproctus (fig. 67).

Thus the traditional distinction between the principal

piece and the end piece, the presence or absence of the

fibrous sheath, is not applicable to avian sperm. What was

believed to be the principal piece by light microscopists is,

23

in fact, the middle piece. The distal end of the flagellum

is smaller in diameter, and the cytoplasmic membrane is

more closely applied to the axoneme. No distinct boundary

between the principal piece and the end piece can be

distinguished. At the very tip of the flagellum the sym-

metry of the axoneme is lost.

The Cytoplasmic Membrane and Annulus

The entire spermatozoon is ensheathed by a plasma mem-

brane originally derived from the spermatid. It tends to

adhere closely over the acrosome and anterior part of the

nucleus. It becomes somewhat slack over the posterior

portion of the nucleus, the neck, and middle piece, and then

adheres progressively tighter as it proceeds down the re-

mainder of the flagellum.

All mammalian sperm have a dense ring or annulus

associated with the cytoplasmic membrane at the caudal end

of the mitochondrial sheath. The annulus of Gallus has

been described by Nagano (1962) and Lake et al. (1968) but

this structure has not been observed in other birds (fig.

36).

A COMPARISON OF AMNIOTE SPERMATOZOA

The fine structure of avian spermatozoa can be compared

with that of other vertebrates to determine which components

they share and which may be unique to birds. The sperm of

fishes and amphibians present such major differences

(Baccetti, 1970) that comparisons with birds are beyond the

scope of this study. Since birds and mammals have reptilian

origins, comparison of the amniotes is more reasonable.

With one exception, spermatozoa of mammals can be dis-

tinguished from those of birds by cursory examination. They

are generally ovate or falciform and flattened dorsoven-

trally. The exception is the echidna, Tachyglossus aculeatus,

whose spermatozoon is quite avian or reptilian in appearance

- a thin, very elongate, fusiform body with a pointed acro-

some (Rothschild, 1962). This prototherian possesses a

number of reptilian characteristics, but the ultrastructure

of its spermatozoon is unknown.

Spermatozoa of reptiles are superficially very similar

to those of birds, so much so they were considered together

as the "sauropsid" type by early microscopists. They both

24

25

possess a cylindrical or fusiform head, frequently with a

pointed acrosome. Their fine structure, however, reveals

that they are quite different. A comparison of bird and

crocodilian sperm would be the most interesting phylogeneti-

cally, because of their common archosaurian origin, but the

morphology of crocodilian sperm is totally unknown.

The various components of bird spermatozoa are compared

below with those of mammals (Fawcett, 1965, 1970; Fawcett

and Phillips, 1970; Hancock, 1966) and reptiles (Hamilton

and Fawcett, 1968; Furieri, 1970).

The Head

In mammals, the acrosomal membranes extend caudally to

cover part of the nucleus, a situation that does not exist

in birds. Some mammals possess an apical body or subacro-

somal material (Hadek, 1969) that may be analogous to the

apical body and associated material described in some birds

(Galliformes, Charadriiformes, and Columbiformes).

In reptiles, a considerable amount of subacrosomal

material occurs (Furieri, 1970). The acrosomal membranes do

not extend over the nucleus, but a slender, tapered exten-

sion of the nucleus projects forward into the subacrosomal

material. This situation resembles that found in the Columbi-

formes except that the projection is identical to the rest

of the nuclear material, not differentiated as in the

26

Columbiformes. The subacrosomal material of reptiles appears

paracrystalline and may include several accessory structures

that are absent in both birds and mammals.

The Tail

All amniotes have a 9+2 axoneme but vary considerably

in the presence and arrangement of other tail components.

All mammalian spermatozoa whose fine structure is

known possess nine accessory or dense outer fibers in the

flagellum (Fawcett, 1970). These arise from the peripheral

doublet tubules but separate from them with maturation,

remaining attached only at their distal extremity. These

fibers assume highly varied and irregular shapes, which

differ among species, and vary considerably in size. They

also tend to terminate distally at different levels of the

axoneme.

Reptiles also have accessory outer fibers but they

remain much smaller and less exaggerated than in mammals.

In turtles, the accessory fibers are uniquely located on

the inner side of the peripheral doublets, within the

axoneme proper. In lizards and snakes they arise from the

outer side of the peripheral doublets and two of them migrate

out to the fibrous sheath, becoming the longitudinal columns

to which the circumferential ribs attach.

27

Among birds, only one order, Passeriformes, is known

to have fully developed accessory fibers. These are all

of equal size and regular shape. They remain attached to

the peripheral doublet and terminate simultaneously along

the distal axoneme. Gallus has very small accessory

fibers, but their presence among other galliform families

or other orders remains to be demonstrated.

Two centrioles are present in reptiles and non-passerine

birds. In passerines only the distal centriole persists in

the mature spermatozoon, but in mammals it is the proximal

centriole. Reptiles possess a neck cylinder which has no

counterpart among birds or mammals (Hamilton and Fawcett,

1968). It encircles the base of the flagellum and the non-

striated connecting piece. All mammals have a striated

connecting piece which fuses with the proximal end of the

accessory fibers. Among birds, only the passerines have a

striated connecting piece and not all species of this order

possess it.

The middle piece of turtle sperm has the greatest

diameter of the gamete and contains globular mitochondria

that have concentric cristae and occasionally dense centers.

Among lizards, the individual mitochondria are not in direct

contact but are separated by, and in some species embedded

within, a dense, opaque material. In snakes the mitochondrial

28

sheath covers two-thirds of the flagellum, and in some

species the mitochondria are slender and convoluted, with

pleomorphic dense bodies interspersed between them.

The mitochondria of mammals form a single or double

helix about the axoneme, remain as separate entities, and

display considerable species variation in size, shape,

number, and pitch of helix (Fawcett, 1962). Among birds,

they may or may not form a helix, and in at least one order

only a single elongate mitochondrion is present. Dense

intermitochondrial bodies are found only in Columbiformes.

The principal piece of reptiles and mammals is enclosed

by a fibrous sheath of circumferential ribs connected by

two longitudinal columns. In reptiles this sheath extends

forward to the base of the neck, and the mitochondria are

concentrically placed external to the sheath. In mammals

the mitochondria are in direct contact with the axoneme,

and the fibrous sheath encloses the axoneme from the distal

end of the middle piece caudally. The juncture of the middle

and principal pieces is marked by a dense annulus in mammals

and reptiles.

Only two avian orders, Galliformes and Tinamiformes,

have a sheath, and it is amorphous rather than a series of

rings. An annulus is present in the fowl but absent among

the other orders studied.

29

Fawcett (1970) has discussed the principal advances of

mammalian spermatozoa over the primitive spermatozoon (9+2

axoneme with a simple ring of mitochondria at the base of

the flagellum, as found in numerous phyla). He considers

them to be: (1) the addition of the accessory fibers to

the axoneme, (2) the enlargement of the middle piece, and

(3) the acquisition of the fibrous sheath of the principal

piece.

Reptilian spermatozoa have made similar advances.

Although the accessory fibers are not fully developed in some

species, all have an enlarged middle piece and a fibrous

sheath.

In birds the middle piece is highly variable in length

but always larger than that of the primitive spermatozoon.

The fibrous sheath occurs only in Galliformes and Tinami-

formes. Accessory fibers are known only in Galliformes and

Passeriformes. The Charadriiformes, Columbiformes, and

Piciformes lack both the fibrous sheath and accessory fibers.

The scarcity of these three characters may indicate that

avian sperm are relatively primitive.

Fawcett (1970) postulates that the lengthened middle

piece of vertebrate spermatozoa has evolved to meet the

energy requirements of the accessory fibers, and both are

correlated with the development of internal fertilization.

30

All avian species have internal fertilization but only two

orders are known to have accessory fibers. All passerines

have accessory fibers but some have a relatively short

middle piece. Many avian orders lack accessory fibers but

have a middle piece equal to or longer than that found

among mammals. The Columbiformes have an extremely

elongate middle piece but no accessory fibers. Therefore,

analysis of the avian middle piece and accessory fibers

would seem to discount this hypothesis.

A COMPARISON OF AVIAN SPERMATOZOA

It was originally hoped that a comparative investi-

gation of avian sperm might shed some light on the evolu-

tionary history of birds. However, as so often befalls

the introduction of a new taxonomic character, be it

morphological, physiological, or biochemical, an examination

of sufficient depth reveals greater variation within a

particular family or order than that which exists between

like taxa. Frequently the most difficult task for a taxono-

mist attempting to define a family or order is to determine

characters common to all members of a group and absent in

all other groups. As the exceptions to the rules become

more numerous, the initial exuberance yields to the reali-

zation that only an arduous, detailed investigation of a

large and almost complete sampling will reveal the secrets

of phylogeny.

This study included examination of sperm from 19 of

the 27 living orders of birds (Wetmore, 1960). A 20th, the

ostrich, has previously been described (Retzius, 1911a).

Those orders which are lacking are comprised of only a few

species each, i.e. the rheas, emus, kiwis, loons, grebes,

penguins, and colies. Most of the species studied (200 of

31

32

281) represent the single order Passeriformes which includes

about 57 per cent of the living species of the world. At

the family level, a more equitable representation has been

attained, with 32 non-passerine and 35 passerine families

included (of 177 living families in the world).

The spermatozoan ultrastructure of only five orders

has been investigated and each is different. It is readily

apparent that the remaining orders need to be similarly

studied. Convergence frequently creates a problem in

deciphering phylogenetic relationships, and sperm morphology

is no exception. One excellent example is already known

(McFarlane, 1963).

The Columbiformes have a unique middle piece which is

readily apparent at low magnifications. Six genera have

been examined (Claravis, Columba, Columbigallina, Leptotila,

Zenaida, Zenaidura) and all possess an extremely elongate

middle piece. The dense intermitochondrial granules and.

nuclear spine are also unknown in other orders.

Galliform spermatozoa have an axonemal sheath, an api-

cal spine, and accessory fibers. The Tinamiformes appear

to possess a similar axonemal sheath (fig. 67) in the prin-

cipal piece but their ultrastructure is unknown. The

illustrations of Retzius (1909) indicate that apical

spines may exist among other birds (perhaps in Anseriformes,

33

Gruiformes, Psittaciformes, Cuculiformes).

The Piciformes seem to have relatively unspecialized

spermatozoa but a number of other orders (Apodiformes,

Strigiformes, Ardeiformes, Pelecaniformes) resemble them in

general appearance and may share some ultrastructural details.

The sperm of the Passeriformes differ from those of all

other birds, and the evolutionary trend of sperm development

in this order seems clear. A detailed analysis of passerine

sperm is beyond the scope of this study, but several gener-

alizations are appropriate. The most conspicuous change in

the head has been the relative size of the acrosome and

nucleus. The acrosome of lower passerines is quite small

but it subsequently comes to dominate the structure of the

head, being several times larger than the nucleus. Simul-

taneously it develops a helical, lateral extension of the

acrosome, a feature that becomes quite exaggerated in some

families. This acrosomal arrangement results in the largest

sperm heads among birds.

Equally significant variations have occurred in the

tail of passerines. It has greatly increased in length,

and concurrently the middle piece has extended farther down

the axoneme. A similar situation exists in the Columbiformes,

which have the tail and middle piece very long, but the

passerines differ in having only a single mitochondrion. The

34

dense outer fibers of the passerine axoneme have also

increased in diameter with the evolution of the longer tail,

and the mechanical attachment of the tail to the nucleus

has altered, although the fine structural details of the

neck are poorly known.

The most distinctive trend among passerines is the

development of their overall helical configuration. The

families of the suborder Tyranni progress from a very gentle

coil of one revolution in the Dendrocolaptidae and Furnari-

idae to a tightly wound corkscrew in Formicariidae, Pipridae,

Cotingidae, and Tyrannidae. The helix is primarily a

feature of the nucleus, with only a minimal helical membrane

on the small acrosome. The middle piece in these families

is very short. The families of the suborder Passeres begin

with the helix primarily restricted to the nucleus, with a

short middle piece, in the Corvidae and Laniidae. Both the

acrosome and middle piece increase in size and helical

development in some families (e.g. Paridae and Vireonidae).

In others (Parulidae, Icteridae, Thraupidae, Fringillidae)

the nucleus shortens, the acrosome lengthens and develops a

wide helical membrane, and the helical mitochondrion extends

almost the full length of the tail.

The functional advantage of this helical configuration

remains obscure. The helical membrane of the acrosome would

35

undoubtedly impart a counterclockwise rotary motion during

the forward progression of the spermatozoon, and this has

been observed in passerines (Nicander, 1970). Thompson

(1966) experimented with models of nudibranch sperm, which

closely resemble those of passerines but have a right-handed

helical membrane attached to the nucleus instead of the

acrosome. He was able to duplicate the rotary motion

observable in living Archidoris sperm.

The low profile of the mitochondrial helix would not

appear to increase the rotation of the spermatozoon to any

great extent, except in the Turdidae where the tail helix

dominates. The function of the mitochondrial helix may be

to reduce the force necessary to propagate wave motion down

the flagellum. Apparently the development of an extremely

long tail in a species utilizing internal fertilization

requires that energy sources remain close to the axoneme for

most of its length. A helical mitochondrion would offer

less resistance than a straight rod or sheath (Andre, in

Baccetti, 1970, p. 272), particularly if the waves are

biplanar. In any event, rotary motion appears to have some

selective advantage, which has resulted in its development

among the most recently evolved families of birds. It has

independently arisen in some members of two other phyla,

Annelida and Mollusca (Franzen, 1970), and in two other

orders of birds.

The Charadriiformes are one of the larger avian orders,

and their sperm are now rather well known. The early light

microscopists (Ballowitz, 1888, 1913; Retzius, 1909)

studied 9 species of this order. I have studied an addi-

tional 21 species from 7 families. The sperm of 6 of these

families (Jacanidae, Charadriidae, Recurvirostridae, Laridae,

Rynchopidae, Alcidae) are remarkably uniform, exhibiting

variation in the relative size of the acrosome and middle

piece but with the same basic morphology (fig. 5).

The sperm of the sandpipers (Scolopacidae) are drama-

tically different from those of the other Charadriiformes

but resemble passerine sperm. The entire head is helical,

as is the middle piece in some species, and the acrosome

has a helical membrane (fig. 71). This configuration is

consistent in the six genera whose sperm is known (Actitis,

Capella, Catoptrophorus, Scolopax, Totanus, Tringa). This

situation is particularly interesting because the sandpipers

and the plovers (Charadriidae) are considered by some taxo-

nomists to be so closely related as to merit only subfamily

distinction, yet those plovers whose sperm is known (Chara-

drius, Vanellus) have sperm typical of the other charadriiform

families (fig. 72). The ultrastructure of neither group is

known, nor have the spermatozoa of intermediate forms, such

as Arenaria, been studied.

A similar but less well documented case is found in

the Procellariiformes. In this instance three families are

known from a single species each. Oceanodroma (Hydrobatidae)

and Diomedea (Diomedeidae) have blunt, rounded acrosomes

(fig. 70), but Puffinus (Procellariidae) has a short and

strongly helical acrosome (fig. 69).

Thus we see an excellent example of convergent evolu-

tion. The appearance of the helical acrosome in only one

of three procellariiform families and one of seven charadri-

iform families can hardly indicate relationship with the

passerines, where we find the full spectrum from slight

to full helix within a single order.

The Passeriformes are thought to be the most recent

order to appear in the evolutionary history of birds. On

the basis of sperm morphology, the passerines display no

relationship to any of the other orders whose sperm is

known, including the Piciformes, which have frequently

been suggested as ancestral to the Passeriformes (Furbringer,

1888).

SUMMARY

Avian spermatozoa exhibit considerable variation in

both gross and fine structure and may be useful indicators

of phylogenetic relationship but knowledge of their ultra-

structure is required. The spermatozoa of 281 species of

birds were surveyed with phase-contrast microscopy. Of

these, 177 species were viewed with the electron microscope

and the ultrastructure of 18 species was studied in detail.

The highly structured fibrous sheath of the principal

piece, found in both reptiles and mammals, is absent in

birds, although two avian orders (Galliformes and Tinami-

formes) have an amorphous sheath. The accessory fibers

of the axoneme, common in reptiles and highly developed

in mammals, also occur in two avian orders (small in

Galliformes but well developed in Passeriformes). On the

other hand, the acrosome of many passerines has achieved a

degree of development unequalled among other vertebrates.

APPENDIX A.

Figures 1 73

KEY TO ABBREVIATIONS

acr acrosome

ap.b. apical body

ap.s. apical spine

a.s. amorphous sheath

ax axoneme

c.c. centriolar complex

c.m. cytoplasmic membrane

c.p. connecting piece

d.c. distal centriole

d.g. dense granule

e.p. end piece

g.s. granular substance

mi mitochondrion

m.p. middle piece

m.s. microtubular spindle

n.m. nuclear membrane

nuc nucleus

o.f. dense outer fiber

p.c. proximal centriole

p.p. principal piece

Figure 1. Diagrammatic representation of the spermatozoon
of the chukar, Alectoris graeca.
(Galliformes, Phasianidae)

A, X2,000; B, X10,000; C I, X50,000

A. Entire spermatozoon. Total length = 90p.

B. Head and middle piece. Length of acrosome = 1.8p,
nucleus = 11.3p, middle piece = 3.7p.

C. Sagittal section of the anterior head. The conical
acrosome has a deep posterior cavity that slips over
the proximal nucleus and encloses the apical spine.
The spine is embedded within the nucleus but is
exterior to the nuclear membrane. A cytoplasmic mem-
brane ensheaths the entire spermatozoon.

D. Sagittal section of the anterior tail. The axoneme
arises from the centriolar complex and is surrounded
by the plate-like mitochondria of the middle piece.

E. Cross section of the head at level e-e, intersecting
the acrosome and apical spine.

F. Cross section of the head at level f-f, intersecting
the acrosome, nucleus, and apical spine.

G. Cross section of the middle piece at level g-g,
intersecting the mitochondrial sheath and the axoneme.

H. Cross section of the principal piece at level h--h,
intersecting the amorphous sheath and axoneme.

I. Cross section of the end piece at level i-i, inter-
secting only the axoneme and cytoplasmic membrane.

acr acrosome
ap.s. apical spine
a.s. amorphous sheath
ax axoneme
c.c. contriolar complex
c.m. cytoplasmic membrane
mi mitochondrion
m.p. middle piece
nuc nucleus
p.p. principal piece

I I

acr

ap.

nuc

mi-

mi-

c.m.

O.s.

axI
ax

E

F

p. P.-

Figures 2 4. Alectoris graeca.

Buffered formalin fixation.

Fig. 2.

Fig. 3.

Fig. 4.

Os04 post-fixation.

Longitudinal section of the nucleus and an axoneme.
Numerous transverse sections the tail, demon-
strating the changes in morphology and reduction
in diameter caudally. Note sections of the
middle piece with surrounding mitochondria,
the principal piece and its amorphous sheath,
and the axoneme of the end piece.
X20,000

Longitudinal section of anterior nucleus and
acrosome. The apical spine is embedded in the
nucleus and extends forward into a cavity
of the acrosome. X75,000

Longitudinal section of the middle piece and
nucleus. The centriolar complex at the base of
the nucleus gives rise to the axoneme which is
surrounded by the mitochondria of the middle
piece. X21,000

acr
ap.s.
ax
cc.c
e.p.
m.p.
nuc
p.p.

- acrosome
- apical spine
- axoneme
- centriolar complex
- end piece
- middle piece
- nucleus
- principal piece

w

. wit

ki. 7?
mb, 'P"

-s C.

S-

wI, t ^u

~- n4.c.

-%. t.

4%

1'..

a.

*Vi

.Ass

,--rr -2r'--- 3*-

alM, l "
Sn-*~
I, ~,"~eat.

ri;"

- F

0. ,

l*

'I

'il

.

41N
a 2,1

4z. zsv

N*

ft

Figure 5. Diagrammatic representation of the spermatozoon
of the sooty tern, Sterna fuscata.
(Charadriiformes, Laridae)

A, X2,000; B, XI0,00; C F, X40,000.

A. Entire sperm. Total length = 60p.

B. Head and middle piece. Length of acrosome = 0.9p,
nucleus = 7.2p, middle piece = 2.2p.

C. Sagittal section of the anterior head. A spherical
apical body is embedded in the nucleus and capped by
the short, conical acrosome. All are encased by the
cytoplasmic membrane.

D. Sagittal section of the anterior tail. The centriolar
complex at the base of the nucleus gives rise to the
axoneme. The mitochondria are loosely arranged about
the axoneme and do not form a helix.

E. Cross section of the proximal principal piece at level
e-e. The axoneme is surrounded by only the cytoplasmic
membrane.

F. Cross section of the distal principal piece at level
f-f. The only change from (e) is a reduction in'
diameter of the axoneme. In the absence of a sheath
or accessory fibers there is no reliable distinction
between a principal piece and an end piece.

acr acrosome
ap.b. apical body
ax axoneme
c.c. centriolar complex
c.m. cytoplasmic membrane
mi mitochondrion
m.p. middle piece
nuc nucleus
p.p. principal piece

B

I

p. p.

*
Q P;

;;

46

- acr

C.c

mi

f f\

E
P."i

F

Sterna fuscata.

Fig,
Fig.

Fig. 6.

Fig. 7.

Fig. 8.

6, 7, 8, & 11 Os04 fixation, uranyl acetate stain.
9 & 10 formalin fixation, whole mount.

Longitudinal section of sperm head. The nucleus
contains numerous cavities randomly scattered
throughout the granular chromatin. The centriolar
complex closely adheres to the base of the nucleus,
and the mitochondria are oblate and loosely
arranged about the axoneme. X27,000

Subtangential section of anterior head. The double
nuclear membrane contrasts with the single membrane
of the acrosome. The plane of section intersects
the edge of the apical body. X42,000

Parasagittal section of sperm head. The apical
body can be seen embedded in the tip of the
nucleus. X54,000

Fig. 9. View of anterior sperm head and apical body
with acrosome detached. X36,000

Fig. 10. View of anterior sperm head with acrosome intact.
X36,000

Fig. 11. Longitudinal section of anterior tail. The
axoneme can be seen emanating from the centriolar
complex. X48,000

acr acrosome
ap.b. apical body
ax axoneme
c.c. centriolar complex
mi mitochondrion
n.m. nuclear membrane
nuc nucleus

Figures 6 11.

.5l
4' "!:" .!
, ..::. ... E: E

4w

w

. ..^3 ,:.
'.-C

OP&. b.

Figure 12. Diagrammatic representation of the spermatozoon
of the rock dove, Columba livia.
(Columbiformes, Columbidae)

A, X2,000; B, X10,000; C F, X50,000

A. Entire sperm. Total length = 160p. The termination
of the mitochondrial sheath of the middle piece is
indicated by an arrow.

B. The head. Length of acrosome = 2.4p, nucleus = 16.Op,
middle piece = 105p.

C. Sagittal section of anterior portion of head. The
apical spine is an extension of the nucleus and is
limited by the double nuclear membrane.

D. Cross section of the head at level d-d, intersecting
the acrosome and apical spine.

E. Sagittal section of anterior tail. The proximal
centriole lies at an angle of 750 to the longitudinal
axis, the distal centriole is parallel to the axis.
The peripheral doublets of the axoneme are continuous
with the triplets of the distal centriole, and the
central tubules arise at the caudal end of the distal
centriole. Dense granules are interspersed among the
mitochondria of the middle piece.

F. Sagittal section of anterior tail, view rotated 90
from (E), demonstrating the triplets of the proximal
centriole.

acr acrosome
ap.s. apical spine
ax axoneme
c.m. cytoplasmic membrane
d.c. distal centriole
d.g. dense granule
mi mitochondrion
m.p. middle piece
nuc nucleus
p.c. proximal centriole

50

A B C

S, --. !
ocr

d-- -': -
':""""' ......* d

*d.g.

m 11
p. -'

nyc

m.p. 0
ax ib- a ;l.
**7/ ^ ^Av-
| _nu ^S

i~: "-I- j 4

Figures 13 15.

Os4 fixation. Lead citrate stain.

Fig. 13.

Fig. 14.

Fig. 15.

Longitudinal section of posterior nucleus,
anterior middle piece, and centriolar
complex. Note the dense granules among the
polygonal mitochondria. X24,000

Longitudinal section of middle piece, demon-
strating the extreme elongation of this com-
ponent and the helical arrangement of the five
rows of polygonal mitochondria surrounding the
axoneme. X11,000

Longitudinal section of the middle piece. The
dimorphic character of the peripheral doublets
and the numerous dense granules (arrows) are
apparent. X39,000

c.c. centriolar complex
m.p. middle piece
nuc nucleus

Columba livia.

52

I P43.

4

or

9-

S.,

e

4

Figures 16 21.

OsO4 fixation. Lead citrate stain.
4

Fig. 16.

Fig. 17.

Fig. 18.

Fig. 19.

Fig. 20.

Fig. 21.

Longitudinal section of posterior nucleus and
centriolar complex. The relative positions of the
proximal and distal centrioles, and the double
nuclear and single cytoplasmic membranes are
apparent (arrows). X48,000

Longitudinal section of the centriolar complex
perpendicular to the axis of the proximal centriole.
Note the dimorphic character of the peripheral
doublets and their continuation into the distal
centriole, the termination of the central tubules
at the caudal end of the centriole, and the triplets
of the proximal centriole. The cristae of the
mitochondria are parallel with the main axis and
the dense granules (arrows) are present right up
to the base of the neck. X52,000

Sagittal section of anterior head. The apical
spine is continuous with the nucleus and within
the double nuclear membrane. It is surrounded
by an electron translucent material and does not
directly contact the acrosome. X68,000

Cross section of the proximal centriole, demon-
strating the triplet character of the centriolar
tubules (arrow). X52,000

Cross section of the middle piece, demonstrating
the five rows of mitochondria and the peripheral
doublets, and central singlets of the axoneme.
X96,000

Cross section of the middle piece. The central
tubules are linked by two arcs, and spokes lead
from the center of the axoneme to each of the
doublets. The five rows of mitochondria and a
dense granule are visible. X110,000

acr
ap. s.
d.c.
d.g.
mi
nuc
p.c.

- acrosome
- apical spine
- distal centriole
- dense granule
- mitochondrion
- nucleus
- proximal centriole

Columba livia.

~2ot

Ii

lq

9

: fil

Figure 22. Diagrammatic representation of the spermatozoon
of the red-bellied woodpecker, Centurus
carolinus (Piciformes, Picidae).

A, X2,000; B, X10,000; C F, X50,000

A. Entire spermatozoon. Total length = 90p.

B. Head and middle piece. Length of acrosome = 1.2p,
nucleus = 13.0p, middle piece = 5.4p.

C. Surface view of acrosome.

D. Sagittal section of acrosome, demonstrating the oblique
juncture of acrosome and nucleus, and the absence of an
apical spine or body.

E. Sagittal section of proximal portion of tail, with deep
implantation fossa of nucleus surrounding the centriolar
complex. Note the closely packed mitochondria.

F. Side view of head-tail juncture. Caudal projections of
the nucleus form a partial hood around the centriolar
complex.

G. Transverse section of the middle piece.

acr acrosome
ax axoneme
c.c. centriolar complex
c.m. cytoplasmic membrane
mi mitochondrion
m.p. middle piece
nuc nucleus
p.p. principal piece

G

m.p.

Centurus carolinus

Serial sections of the middle piece of four spermatozoa.
Os4 fixation. Uranyl acetate stain.

Fig. 23.

Fig. 24.

Longitudinal section. The undulation of the
spermatozoa through the plane of section results
in a median section of the axoneme and a
subtangential section of the nucleus.
X30,000

Subtangential section of the middle piece and
principal piece. Note the tightly packed
polygonal mitochondria and the implantation
fossa of the nucleus surrounding the centriolar
complex. X30,000

ax axoneme
c.c. centriolar complex
m.p. middle piece
nuc nucleus
p.p. principal piece

Figures 23 and 24.

23. 24.

V

i

." : -'. ..
-C *

, ., Ji

tPt
CtCt

Erf .p
r;s~ "'1~ "" *" C.

p..

Y I .

I -

Figures 25 31.

Os4 fixation.

Fig. 25.

Fig. 26.

Fig. 27.

Fig. 28.

Fig. 29.

Fig. 30.

Fig. 31.

Centurus carolinus

Uranyl acetate stain.

Longitudinal section of the spermatozoon head.
Note the junction of the nucleus and the
acrosome, and the acrosomal vacuole (arrow).
X35,000

Longitudinal section of the spermatozoon head.
Note the nucleus-acrosome junction and the double
nuclear membrane (arrow).. X35,000

Cross section of middle piece, acrosome, and
distal axoneme. Note the three membrane-limited
vacuoles of the acrosome (arrow). X35,000

Oblique section of the nucleus-acrosome juncture.
X61,000

Oblique section of head-tail juncture. The
centriolar complex closely adheres to the
implantation fossa of the nucleus. X45,000

Oblique section through the centriolar complex,
surrounded by the hood-like extensions of the
nucleus. X58,000

Oblique section of the nucleus-tail juncture.
Compare with fig. 29 and 30. The nucleus
extends farther posteriorly in this aspect,
demonstrating the asymmetrical hood. The double
nuclear and single cytoplasmic membranes are
apparent (arrow). X42,000

acr acrosome
ax axoneme
c.c. centriolar complex
m.p. middle piece
nuc nucleus

T .4

S.l. I ~

I r

H i'

Figure 32. Diagrammatic representation of the spermatozoon
of the great crested flycatcher, Myiarchus
crinitus (Passeriformes, Tyrannidae).

A, X2,000; B, X10,000; C J, X50,000

A. Entire spermatozoon. Total length = 50p.

B. Head and middle piece. Length of acrosome = 2.5u,
nucleus = 18.5u, middle piece = 3.3p. The entire
nucleus is helical, with the helix continued anteriorly
by the helical membrane of the acrosome, and posteriorly
by the single helical mitochondrion around the axoneme.

C. Sagittal section of the acrosome-nucleus juncture,
which is slightly oblique.

D. Longitudinal section of the nucleus-tail juncture.
The nucleus has only a shallow implantation fossa to
receive the centriolar complex. The single mitochondrion
is displaced laterally and completes one helical
revolution.

E. Sagittal section of the caudal extremity of the middle
piece. There is no annulus around the axoneme.

F. Cross section of the acrosome at level f-f.

G. Cross section of the acrosome-nucleus juncture at
level g-g.

H. Cross section of the middle piece at level h-h. Note
the asymmetry of the 9+9+2 axoneme and the laterally
displaced mitochondrion.

I. Cross section of the principal piece at level i--i.

J. Cross section of the end piece at level j--j. The
dense outer fibers of the axoneme are not present.

acr acrosome
ax axoneme
c.c. centriolar complex
mi mitochondrion
m.p. middle piece
nuc nucleus

A B

acr

I
'a..

J
C

h-

m.p.

ax

Figures 33 36.

Os04 fixation.

Fig. 33.

Fig. 34.

Fig. 35.

Fig. 36.

Myiarchus crinitus

Lead citrate stain.

Entire head, whole mount. Arrow denotes the
juncture of the middle piece and nucleus.
X8,000

Longitudinal sections of nucleus, the nucleus-
acrosome juncture, and acrosome tip (note double
membrane at arrow). Cross sections of acrosome
and acrosome tip, and numerous cross sections of
principal piece. The very dense, irregular areas
are precipitated lead. X38,000

Longitudinal section of nucleus-tail juncture.
X36,000

Sagittal section of distal end of the middle piece.
There is no annulus posterior to the mitochondrion.
The dense outer fibers are in intimate contact
with the peripheral tubules of the axoneme.
X37,000

acr acrosome
mi mitochondrion
nuc nucleus
p.p. principal piece

S3.

64

F-1

.il

tff

1

/ *I

.

Figures 37 40. Western kingbird, Tyrannus verticalis
(Passeriformes, Tyrannidae)

Formalin fixation. Os4 post-fixation (fig. 38-40).

Fig. 37.

Entire spermatozoon. Whole mount. Arrows denote
the anterior and posterior limits of the nucleus.
X5,200

Fig. 38. Longitudinal section of anterior middle piece.
Note the asymmetrical mitochondrion. X32,000

Fig. 39. Oblique section of anterior middle piece.
X47,000

Fig. 40. Cross section of middle piece. Note the asym-
metry of the mitochondrion and the 9+9+2 axoneme.
X145,000

ax axoneme
mi mitochondrion
nuc nucleus

31.

I

'YICI~ :C

Figure 41. Diagrammatic representation of the spermatozoon
of the violet-green swallow, Tachycineta
thalassina (Passeriformes, Hirundinidae)

A, X2,000; B, X10,000; C J, X25,000

A. Entire spermatozoon. Total length = 285p.

B. Head. Length of acrosome = 13.5p, nucleus = 4.5p.

C. Sagittal section of acrosome-nucleus juncture.

D. Sagittal section of anterior tail. The dense outer
fibers are continuous with the connecting piece at the
base of the nucleus. The helical mitochondrion has a
lateral extension which curves around the axoneme.
One or two small spherical mitochondria may also be
present. The granular substance surrounds the mito-
chondrion and the axoneme and extends a short distance
posteriorly.

E. Sagittal section of the middle piece at the posterior
termination of the granular substance. The helical
mitochondrion has a triangular cross section at this
point.

F. Cross section of the acrosome at level f-f.

G. Cross section of the acrosome at level g-g.

H. Cross section of the middle piece at level h--h, through
the lateral extension of the helical mitochondrion.
Note the 9+9+2 axoneme.

I. Cross section of the mitochondrion at level i-i,
intersecting the granular substance and the triangular
helical mitochondrion.

J. Cross section of the middle piece at level j--j. Note
the asymmetry of the helical mitochondrion and the 9+9+2
axoneme.

acr acrosome
c.p. connecting piece
g.s. granular substance
mi mitochondrion
m.p. middle piece
nuc nucleus
o.f. dense outer fibers

F

nuc

nuc

Tachycineta thalassina.

Formalin fixation. Fig. 42 & 45 stained with uranyl
acetate. Fig. 43 & 44 shadowed with chromium.

Fig. 42.

Longitudinal section of spermatozoon head. Note
the dense core area of the acrosome, the less
dense surrounding material, and the dense cross
sections of the helical membrane. In the lower
right-hand corner is a cross section of the
anterior middle piece, distal to the granular
substance. Note the triangular mitochondrion.
X13,000

Fig. 43.

Fig. 44.

Fig. 45.

Whole head, shadowed with chromium.
denotes the direction of shadowing.

The arrow
X8,000

A portion of the middle piece, shadowed with
chromium. The helical configuration of the
mitochondrion and the individual dense outer
fibers are apparent. X26,000

A longitudinal section of the posterior acrosome
and its junction with the nucleus (lower) and a
cross section of the acrosome (upper). Note the
dense core of the acrosome and the less dense
material at the base of the helical membrane.
X23,000

acr acrosome
mi mitochondrion
m.p. middle piece
nuc nucleus
o.f. dense outer fibers

Figures 42 45.

70

P ,

u

;*0
V

45.
,,

".& -. .

; .1
2i

tL. 4

S
a

A 2.

Figures 46 50. Tac

Formalin fixation.

Fig. 46.

Fig. 47.

Fig. 48.

Fig. 49.

Fig. 50.

hycineta thalassina.

Uranyl acetate stain.

Longitudinal section of the anterior middle
piece. The dense outer fibers continue into
the connecting piece at the base of the nucleus.
The granular substance surrounds the mitochondria
and the axoneme. X30,000

Longitudinal section of the anterior middle
piece. Note the short length of the granular
substance which extends from the neck down the
middle piece for 4u. X18,000

Cross section of the anterior middle piece,
immediately posterior to the neck. Note the
large mitochondrion with a lateral extension
encircling the axoneme through an arc of 2700,
and the smaller spherical mitochondrion at the
same level. X38,000

Cross section of the middle piece. Note the
dense outer fibers attached to the peripheral
doublets of the axoneme, and the triangular
mitochondrion. X70,000

Oblique section of the nucleus-tail juncture,
and two cross sections of the posterior middle
piece, where the elongate mitochondrion has
assumed a cylindrical configuration. X33,000

c.p. connecting piece
g.s. granular substance
mi mitochondrion
m.p. middle piece
o.f. dense outer fibers

S. --- -- at .: a

.^Iif
*A *B" -^

'gad

;46.
a ^jf

-w

Sc'. _
a-.

49.

0

Figure 51. Diagrammatic representation of the spermatozoon
of the tufted titmouse, Parus bicolor
(Passeriformes, Paridae) and the red-eyed vireo,
Vireo olivaceus (Passeriformes, Vireonidae).

A & B, X15,000; C & D, X8,300

A. Parus bicolor. Head. Total length of spermatozoon = 90p,
acrosome = 7.11, nucleus = 5.8p, middle piece = 50p.
The nucleus is helical and has a secondary helical
constriction which forms a slight furrow along its length.

B. Parus bicolor. Sagittal section of A.

C. Vireo olivaceus. Head and middle piece. Total length
of a spermatozoon = 80p, acrosome = 8.0p, nucleus =
5.5p, middle piece = 10.p. The middle piece extends
for only one-sixth of the 66p tail.

D. Vireo olivaceus. Sagittal section of C.

acr acrosome
m.p. middle piece
nuc nucleus

acr

nuc

m.

Parus bicolor and Vireo olivaceus.

Fig. 52.

Fig. 53.

Fig. 54.

Fig. 55.

Parus bicolor. Sagittal section of head and
middle piece. The granular chromatin contains
numerous cavities and has a secondary helical
constriction which runs the full length of the
nucleus. A deep implantation fossa encompasses
the centriolar complex. Glutaraldehyde fixation.
Uranyl acetate stain. X17,000

Parus bicolor. Head, whole mount.
Formalin fixation.

X9,300

Vireo olivaceus. Head and middle piece, whole
mount. The anterior and posterior limits of
the middle piece are indicated by the arrows.
Formalin fixation. X4,800

Vireo olivaceus. Sagittal section of nucleus and
middle piece. The axoneme is tangentially
attached to the base of the nucleus. Glutaralde-
hyde fixation. Uranyl acetate stain. X23,000

acr acrosome
ax axoneme
c.c. centriolar complex
mi mitochondrion
nuc nucleus
p.p. principal piece

Figures 52 55.

76

s6. .rr.
5

*.

VAA V
C.C.

-Va

"I
to

'In, i, p... Ii

[ k' i

Figure 56.

Diagrammatic representation of the spermatozoa
of the robin, Turdus migratorius (Passeriformes,
Turdidae) and the summer tanager, Piranga rubra
(Passeriformes, Thraupidae).

A & B, X15,000; C & D, X13,000

A. Turdus migratorius. Head. Total length of spermatozoon
= 70p, acrosome = 6.7p, nucleus = 2.5p, middle piece
= 46p. This family is characterized by a very large
helical mitochondrion, resulting in the mitochondrial
helix having a larger diameter than the helical membrane
of the acrosome.

B. Turdus migratorius.
middle piece.

Sagittal section of head and anterior

C. Piranga rubra. Head. Total length of spermatozoon =
170p, acrosome = 12.0p, nucleus = 3.Op, middle piece =
146p. This family is characterized by an extremely
wide helical membrane on the acrosome. Narrow exten-
sions of the acrosome extend to the edge of the helix.

D. Piranga rubra.

Sagittal section of head.

acr acrosome
m.p. middle piece
nuc nucleus

*acr

nuc-

nuc

m.

Figures 57 59.

Fig. 57.

Fig. 58.

Fig. 59.

Turdus migratorius. Head and middle piece.
Whole mount. The arrows mark the anterior and
posterior limits of the middle piece, which
extends for most of the axoneme length.
Formalin fixation. X8,200

Piranga rubra. Head, demonstrating the broad
helical membrane of the acrosome.
Formalin fixation. X7,600

Tangara gyrola, the bay-headed tanager (Passeri-
formes, Thraupidae). Entire spermatozoon.
Total length 148p. Arrows denote the anterior
and posterior limits of the middle piece.
Formalin fixation. X2,900

a

Figure 60.

Testicular section of developing spermatids
of Piranga rubra. A striated connecting
piece can be observed at the posterior end
of one nucleus. The acrosome contains material
of differing densities, as does the proximal
middle piece, where the granular substance is
contrasted with the microtubular spindle.
Compare the large, elongate mitochondrion of
the spermatid with the numerous, smaller
typical mitochondria of the surrounding Sertoli
cells.
OsO4 fixation. X20,000

acr acrosome
c.p. connecting piece
g.s. granular substance
mi mitochondrion
m.s. microtubular spindle

*.

'4
I

I 11

A k, 2A

F ~rT t~e 4 PYA

Figures 61 63.

Testicular sections of developing sperma-
tids of Piranga rubra.

Os04 fixation.

Fig. 61.

Fig. 62.

Fig. 63.

Transverse sections of acrosomes. Some shi
three distinct density areas, with dense
material at the apex (arrows) while other
sections, apparently more anterior, have a
uniform medium density. Some of the acro-
somes exhibit a definite double limiting
membrane. X1

ow

6,000

Cross sections of the middle piece. Three
distance structures are present: the 9+9+2
axoneme, the helical mitochondrion, and the
microtubular spindle, which can be seen to
include cisternae of endoplasmic reticulum,
both open and collapsed. X60,000

Cross sections of the middle piece. Isolated
membranes from the collapsed cisternae can be
seen within the microtubular spindles. X30,000

mi mitochondrion
m.s. microtubular spindle

I v).

A i

w6

1

*

2p*i
444. .~

h'

*4 R

.
* *9^

JC

- 4.

.a~

i
r ~

v r
L:
I

r!

si.

S"9

.eq *

..
1

I
s~
L s

~b":

Figures 64 & 65.

Os04 fixation.

Fig. 64.

Fig. 65.

Testicular sections of spermatids of
the rufous-sided towhee, Pipilo
erythrophthalmus.
(Passeriformes, Fringillidae)

Uranyl acetate stain.

Microtubular spindles can be seen adjacent to
the acrosome-nucleus juncture at far left and
far right, and the nucleus at center right.
Cisternae of the endoplasmic reticulum have been
enclosed within the spindle at far left. The
double membrane of the acrosome is readily
apparent (arrow). The transverse section of the
nucleus (center right) suggests that the micro-
tubules may actually be within the perinuclear
cisterna. X27,000

The microtubular spindle extends caudad to the
middle piece, where it coils about the axoneme.
The collapsed cisternae of endoplasmic reticulum
appear as random membranes within the bundles
of tubules. Compare the typical mitochondria
of the Sertoli cells with the elongate helical
mitochondrion of the spermatid, whose cristae
have not completed their reorientation to the
longitudinal axis. X14,000

acr acrosome
mi mitochondrion
m.s. microtubular spindle
nuc nucleus

86

''1.1 ,
"r ;i |

r?

o-e

00,

Ilkt
.....

14%

... .
.,i E `Z

,lo.
PsI "b '"~ '
'' .' sl!-

B'yC e r' .. .~~ i

,, .:

Some representative spermatozoa.

Formalin fixation.

Whole mounts

Fig. 66. The head of the Tataupa tinamou, Crypturellus
tataupa (Tinamiformes, Tinamidae). X1l,000

Fig. 67. Crypturellus tataupa. A break in the sheath of
the principal piece reveals the axoneme and the
thickness of the sheath. X27,000

Fig. 68. Crypturellus tataupa. The acrosome has a spade-
like anterior projection. X76,000

Fig. 69. The head of the wedge-tailed shearwater, Puffinus
pacificus (Procellariiformes, Procellariidae),
demonstrating the helical acrosome. X13,000

Fig. 70.

Fig. 71.

Fig. 72.

Fig. 73.

The head of the Laysan albatross, Diomedea
immutabilis, which lacks the helical acrosome.
X9,000

The head of the common snipe, Capella gallinago
(Charadriiformes, Scolopacidae), with helical
acrosome and nucleus. X9,000

The head of the killdeer, Charadrius vociferous
(Charadriiformes, Charadriidae) with cylindrical
nucleus and blunt acrosome. X16,000

The head and tail of the three-wattled bellbird,
Procnias tricarunculata (Passeriformes, Cotingidae)
which is representative of the most extreme helical
configuration among the suborder Tyranni.
X9,000

acr acrosome
m.p. middle piece
nuc nucleus
p.p. principal piece

Figures 66-73.

CLcr

IuC

&cr

73.

mmh,,

-- iii* *i 'ii ilI: i

"* ::::'i, l l
^I'.'"'2l
:'*::i

^ ^

Acr

hueA

APPENDIX B.

CLASSIFICATION AND LIST OF SPECIES STUDIED

Orders and Families follow Wetmore, 1960.
Asterisk (*) indicates ultrastructure studied.

Tinamiformes
Tinamidae

Nothoprocta cinerascens
N. pentlandii

Procellariiformes
Diomedeidae
Diomedea immutabilis
Procellariidae
Puffinus pacificus
Hydrobatidae
Oceanodroma leucorhoa

Pelecaniformes
Phaethontidae
Phaethon rubricauda
Sulidae
Sula dactylatra
S. leucogaster
Anh ingidae
Anhinga anhinga
Fregatidae
Fregata minor

Ciconiiformes
Ardeidae
Ardea herodias
Butorides virescens
Ixobrychus exilis
Threskiornithidae
Eudocimus albus

Anseriformes
Anatidae
Somateria mollissima

Crypturellus tataupa

S. sula

Bubulcus ibis
Casmerodius albus

Falconiformes
Cathartidae
Cathartes aura
Accipitridae
Buteo lineatus

Galliformes
Tetraonidae
Bonasa umbellus
Phasianidae
*Alectoris graeca
Callipepla squamata

Gruiformes
Rallidae
Fulica americana
Rallus longirostris

Charadriiformes
Jacanidae
Jacana spinosa
Charadriidae
Charadrius vociferus
Scolopacidae
Actitis macularia
Catoptrophorus semipalmatus
Recurvirostridae
Himantopus mexicanus
Laridae
Larus atricilla
L. argentatus
L. marinus
Gelochelidon nilotica
Anous minutus
A. stolidus
Rynchopidae
Rynchops nigra
Alcidae
Fratercula nigra

Columbiformes
Columbidae
*Columba livia
C. nigrirostris
Columbigallina passerina
C. talpacoti

B. platypterus

Colinus virginianus
Gallus gallus

Gallinula chloropus

C. wilsonia

Capella gallinago

Sterna albifrons
*S. fuscata
S. hirundo
S. lunata
S. paradisaea
Gygis alba

Claravis pretiosa
Leptotila verreauxi
Zenaida asiatica
Zenaidura macroura

	Title Page
	Acknowledgement
	Table of Contents
	List of Figures
	Abstract
	Introduction
	Methods
	Ultrastructure
	A comparison of amniote sperma...
	A comparison of avian spermato...
	Summary
	Appendix A: Figures 1-73
	Appendix B: Classification and...
	Literature cited
	Biographical sketch

MILLISECOND	CLASS.METHOD	MESSAGE
0	sobekcm_page_globals.constructor
0	sobekcm_page_globals.constructor	Application State validated or built
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.constructor	Navigation Object created from URI query string
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.display_item	Retrieving item or group information
0	sobekcm_page_globals.get_entire_collection_hierarchy	Retrieving hierarchy information
0	sobekcm_assistant.get_entire_collection_hierarchy
0	cached_data_manager.retrieve_item_aggregation
0	cached_data_manager.retrieve_item_aggregation	Found item aggregation on local cache
0	item_aggregation_builder.get_item_aggregation	Found 'all' item aggregation in cache
0	system.web.ui.page.page_load (ufdc.page_load)
0	sobekcm_page_globals.constructor.on_page_load
0	html_echo_mainwriter.add_style_references	Adding style references to HTML
0	html_echo_mainwriter.add_text_to_page	Reading the text from the file and echoing back to the output stream
36	html_echo_mainwriter.add_text_to_page	Finished reading and writing the file