[ 490 ]

THE DEVELOPMENT OF THE HUMAN VAGINA

BY D. BULMER
Anatomy Department, University of Aberdeen
The literature on the development of the human vagina is already so abundant that
this further contribution is made only with some trepidation. However, the findings
which form the substance of this paper seem of sufficient significance to be recorded,
since they clearly demonstrate a method of vaginal development which has not so
far been generally accepted in this country. The material studied, while not always
forming as complete a series as might be wished, has in most cases been in excellent
condition for detailed histological examination, and it is possible that the previous
confusions and disagreements on this subject were largely due to deficiencies in the
material available to earlier workers.
Bloomfield & Frazer (1927) pointed out that the descriptions of vaginal develop-
ment in the literature could be divided into three main groups, depending on the
structures which their authors believed to take part in the formation of the vagina,
and it will be convenient to employ this classification here.
In the first group are those accounts which have attributed the origin of the
entire vagina to the lower end of the Mtillerian utero-vaginal canal. This opinion first
gained general acceptance from the demonstration by Thiersch and others, work
which is well summarized by Banks (1864), of the fusion of the Mullerian ducts in
the genital cord, and in this century has been supported by Felix (1912), Bloomfield
& Frazer (1927), Hunter (1930) and von Lippmann (1939). It may be regarded as
the 'classical' account of vaginal development, and was for many years the standard
textbook description.
The second group consists of those descriptions which supposed that the vaginal
epithelium was formed in part (Tourneux & Legay, 1884; Mijsberg, 1924) or in
whole (Hart, 1896, 1901 and 1911; Kempermann, 1931) from the lower ends of the
Wolffian ducts.
The third group includes the accounts of all those workers who believed that the
epithelium of the urogenital sinus made a contribution to the vagina. Before the
time of Thiersch it was accepted that the entire vagina was derived from the sinus
(Muller, 1830; Rathke, 1832; Valentin, 1835), but this opinion was discarded when
the fusion of the Mullerian ducts became known. Retterer (1891), however,
suggested that while most of the vagina was formed from the utero-vaginal canal
a short lower segment arose by the splitting of the upper part of the sinus into
dorsal and ventral channels, and this view was later supported by Bolk (1907).
Koff (1933), whose work is recognized by most current British and American
textbooks, believed that the lower one-fifth of the vagina was derived from the
sinus by the growth of paired epithelial 'sino-vaginal bulbs' from the dorsal sinus
wall, these bulbs fusing together to form a lower vaginal segment. On the other
hand, Spuler (1930), Vilas (1932), Kempermann (1935) and Meyer (1934-38)
 The development of the human vagina4 491
believed that a proliferation of cells from the dorsal wall of the sinus gave rise to the
epithelial lining of the entire vagina, displacing the epithelium of the utero-vaginal
canal as far as the cervix. This opinion has recently been supported by Politzer
(1955), and Zuckerman (1940) showed that there was a considerable body of endo-
crinological evidence in its favour.
MATERIAL
The material consists of thirteen embryos and foetuses, ranging from 28 to 375 mm.
crown-rump length, and an infant of 1 month. The majority of the specimens were
already sectioned and stained with haematoxylin and eosin, alternately with
haematoxylin and eosin and a trichrome stain, or by the de Castro method of silver
impregnation. The remainder were sectioned at 6-10 /c, either transversely or
coronally, and the sections stained mainly with haematoxylin and eosin or a tri-
chrome stain. With some of the larger specimens slides were also stained with
Best's ammoniacal carmine and by the periodic acid-Schiff technique (Gomori,
1952).
28 mm. embryo
In the lower part of the genital cord the Mullerian ducts are in apposition with
each other, their medial walls forming a septum between the two lumina. They are
lined by a closely packed columnar epithelium and lie between the two Wolffian
ducts, rather larger structures with a cubical epithelium. The caudal ends of the
Mullerian ducts do not quite reach the dorsal wall of the sinus, remaining separated
from the Mullerian tubercle by a small mass of mesoderm, but on either side they
are in close contact with the lower ends of the Wolffian ducts (PI. 1, fig. 1).
Most of the pars pelvina of the sinus is lined by an epithelium consisting of three
or four layers of small, darkly staining cells with closely arranged nuclei, but in the
region of the Mullerian tubercle the epithelium of the dorsal sinus wall is differentiated
into a thin basal layer of deeply staining cells overlain by one to three layers of
larger and very pale staining cells with relatively smaller nuclei. At the Wolffian
openings, on either side of the Mullerian tubercle, these pale cells seem to be com-
pressed together, and they extend backwards for a short distance into the lower ends
of the Wolffian ducts themselves.
42 mm. foetus
The lower portions of the Mullerian ducts are completely fused together as the
utero-vaginal canal and the caudal tip of the Mullerian epithelium forms a solid
mass in contact with the dorsal wall of the sinus, though a small central mass of
mesoderm interrupts this contact in the midline. On either side of the Millerian
tubercle the Wolffian ducts enter the sinus, and are now of about the same calibre
as the utero-vaginal canal. The sinus epithelium presents the same features as in
the 28 mm. specimen (PI. 1, fig. 2).
50 mm. foetus
This specimen shows no significant developmental change, except that the sinus
epithelium in the region of the Miillerian tubercle is now considerably thicker. This
epithelium, divided into basal deeply staining and superficial pale-staining zones,
will in future be referred to as the differentiated type of sinus epithelium.
492 D. Bulmer
65 mm. foetus
The Mullerian utero-vaginal canal is a large structure, oval in cross-section and
with its long axis lying transversely (P1. 1, fig. 3). Throughout most of its extent it is
lined by a closely packed columnar epithelium, but about 03 mm. above the caudal
end the lining cells lose their columnar arrangement and form a thick stratified
polygonal epithelium, which at the lowermost tip completely occludes the lumen.
The Wolffian ducts are very small structures, lined by a cubical epithelium, and lie
on either side of the utero-vaginal canal. There is a short segment where the
Wolffian ducts have completely disappeared, just above their lower ends, but they
reappear below this to join the dorsal wall of the sinus.
The urogenital sinus shows a very significant advance in the region previously
occupied by the Mullerian tubercle (P1. 1, figs. 4, 5, 6). On either side the Wolffian
ducts join 'dorso-lateral projections' of the sinus, solid epithelial structures which
prevent the communication of the ducts with the sinus lumen and are in contact on
their dorso-medial aspects with the solid mass of Mullerian epithelium at the caudal
end of the utero-vaginal canal. More medially, the Mullerian epithelium is displaced
dorsally by a proliferation of darkly staining cells from the dorsal sinus wall,
between the bases of the two dorso-lateral projections (PI. 1, fig. 4). The dorso-
lateral projections are each formed by a central mass of pale-staining cells, with
small nuclei, surrounded by a thin basal layer of more darkly staining cells, and the
cubical cell lining of the lower ends of the Wolffian ducts is applied to their dorsal
aspects. Caudal and ventral to the proliferation of darkly staining cells from the
dorsal sinus wall, the pale-staining cells extend forwards and medially to form a
crest projecting into the dorsal aspect of the sinus lumen (PI. 1, figs. 5, 6)-the
structure identified by Kempermann (1931) as the Wolffsche Kamm. Apart from
this differentiation in the region previously occupied by the Mullerian tubercle, the
remainder of the pars pelvina is lined by a darkly staining stratified epithelium of
six or seven layers of small cells.
In this specimen, therefore, the sinus cells in the region which was previously
lined by the differentiated type of epithelium form a proliferative zone, apparently
composed of three elements. On either side are the dorso-lateral projections,
associated with the Wolffian openings, while between them is a proliferation of
darkly staining cells from the dorsal sinus wall.
68 mm. foetus
This specimen is at only a slightly later developmental stage than the 65 mm.
foetus, but several interesting differences may be noted. The Wolffian ducts are
present throughout the whole of the genital cord as small, but easily recognizable
structures. The occlusion of the lumen of the caudal portion of the utero-vaginal
canal reaches further cranially than in the 65 mm. specimen, to a distance of about
0O8 mm. above the junction of Mullerian and sinus epithelia. A small mesodermal
mass is enclosed in the root of the proliferation of darkly staining cells from the
dorsal wall of the sinus, splitting it into bilateral elements which meet each other
behind the mesodermal mass and are there in contact with the caudal end of the
Mullerian epithelium (Text-fig. 1 a).
 The development of the human vagina 493
The most interesting feature is the appearance of a characteristic system of
longitudinal folds in the sinus wall (Text-fig. 1), foreshadowed at the 65 mm. stage
but now very much more distinct. It may be noted that there is a pronounced
fold-termed the 'urethro-vaginal fold '-demarcating from the rest of the sinus its
cranial and dorsal portion which forms the zone of contact with the Mullerian tissue.
The dorsal margins of the urethro-vaginal folds run from the dorsal wall of the sinus,
below the dorso-lateral projections, to meet each other behind the lower end of the
urethra, and limit the area, dorsal and cranial to them, which is occupied by the

dp Ct
va

a b cd
b dl
d vX

I VfVf X ase

Text-fig. 1. 68 mm. foetus. The figure on the left is a graphic reconstruction (x 20) of a median
sagittal section through the lower end of the utero-vaginal canal and the urogenital sinus.
The figures on the right, a-g, are transverse sections at the levels indicated in the sagittal
section. The dorsal aspect of the sections is towards the left. The Mullerian epithelium is
shown by the solid black shading, and the sinus epithelium by the stippling. A small area of
connective tissue, ct in fig. 1 a, interrupts the sinus proliferation in the midline. The folds of
the sinus wall are indicated in the transverse sections, and their relative positions in the
sagittal section are shown by the dotted lines. dl, dorso-lateral fold; md, median dorsal fold;
1, lateral fold; vi, ventro-lateral fold; uv, urethro-lateral fold; v, median ventral fold. The
dorso-lateral sinus bay lies between the lateral and dorso-lateral folds (figs. If, Ig) or between
the lateral and median dorsal folds (fig. le), and is joined at ba by the duct of Bartholin's
gland. The ventro-lateral bay is between the lateral and ventro-lateral folds. It will be
noticed that the relief of the sinus wall compares very closely with that described by Mijsberg
(1924) and Politzer (1952) at similar developmental stages.

differentiated type of sinus epithelium. The pattern of the folds in the wall of the
human sinus has been described previously by Mijsberg (1924) and Politzer (1952).
They are indicated in the diagrams in Text-figs. 1-4, and a more detailed account
has been given elsewhere (Bulmer, 1955).
14 week foetus
The dorso-lateral projections can be identified, similar in structure to those of
the 65 mm. stage, and their dorso-medial aspects are in contact with the Mullerian
epithelium. Between their bases is the mass of darkly staining cells proliferating
from the dorsal wall of the sinus, ventral to the Mullerian epithelium and now more
extensive than in the 65 mm. foetus.
494 D. Bulmer
94 mm. foetus
There is a considerable increase in size of the sinus compared with the 68 mm.
stage (Text-fig. 2), but the arrangement of the longitudinal folds remains essentially
unchanged. The urethro-vaginal fold is a marked feature, accentuated by the
enlargement of the sinus which lies dorsal and cranial to it (Text-fig. 2b). This
enlargement is associated with a considerable thickening of the epithelium, which
now consists of a basal zone of five or six layers of small, darkly staining cells, with
relatively large nuclei, clearly demarcated from a superficial zone of three or four
layers of larger, clear staining polygonal cells with relatively much smaller nuclei.
There is a striking distinction between the two zones in the low-power view (PI. 2,
fig. 7), and this epithelium is markedly different from the much thinner, un-
differentiated type of epithelium, of four or five layers of smaller, rather darkly
staining cells, which lines the rest of the pars pelvina.

\ \ a b C d e

d d/ dI. =Jzd j
t3 dl
f g h

Text-fig. 2. 94 mm. foetus. Median sagittal section through the sinus and the lower end of the
vagina, with the corresponding coronal sections ( x 13k). Shading and lettering as for Text-
fig. 1. The dotted lines in the sagittal section indicate the relative positions of the folds in the
upper part of the sinus wall. The coronal sections, a-i, are shown with their cranial ends
(i.e. the ventral wall of the sinus) towards the top.

The enlargement of the sinus behind the urethro-vaginal folds also forms the root
of a short 'sinus upgrowth', projecting dorsally and cranially from the sinus to
meet the caudal end of the Mullerian epithelium. The lower end of the sinus up-
growth contains paired lumina, continuous with the sinus lumen, and the epithelium
is of the same differentiated type as that of the sinus enlargement. Followed further
cranially the lumina disappear, and the sinus upgrowth forms a solid crescentic mass
of epithelium in which the basal and superficial zones are still distinct (PI. 2, fig. 8),
though the basal cells are not so prominent as they are more caudally. In addition,
a small mass of darkly staining cells, the caudal end of the Mullerian tissue, is
embedded in this cranial end of the sinus upgrowth.
As the vaginal mass is followed still further cranially the left side of the sinus
upgrowth is joined by a short persistent segment of the lower end of the left Wolffian
duct, and the Mullerian epithelium comes to occupy a gradually increasing area in
 The development of the human vagina 495
the centre. Eventually, about 240 ic above the root of the sinus upgrowth, the sinus
cells are completely replaced by the solid Mullerian epithelium, which is canalized
a short distance above this as the lower end of the utero-vaginal canal. This part
of the canal is lined by a three- or four-layered stratified polygonal epithelium,
which extends cranially to meet the columnar cell lining of the upper part. The
level of junction between these two types of Mullerian epithelium coincides with
a fusiform swelling of the genital cord, corresponding with that identified by Koff
(1933) as the site of the future cervix.
It can be appreciated that several changes have occurred to reach this stage of
development. The three components which formed the proliferation of the sinus
from the 65 mm. stage onwards-the two dorso-lateral projections and the darkly
staining cells between them-have apparently fused together to form a single mass
which extends dorsally and cranially as the sinus upgrowth. The lower ends of the
Wolffian ducts have disappeared except for a short persistent segment on the left
side, which, as might be expected, joins the side of the sinus upgrowth. In addition,
the 'vaginal' portion of the utero-vaginal canal, so far as this can be defined, is now
entirely lined by a stratified polygonal epithelium, presumably derived from the
original columnar Mullerian epithelium. One of the most interesting features,
however, is the differentiation of the sinus epithelium behind and above the urethro-
vaginal folds. A similar differentiation, though less marked, has been noted since
the 28 mm. stage, always confined to this particular region of the sinus, and it is
from this differentiated type of epithelium that the sinus upgrowth appears to
arise.
112 mm.foetus
The configuration of the urogenital sinus shows little change from the 94 mm.
stage, though the pars phallica is now becoming more dorso-ventrally elongated.
The sinus upgrowth extends for about 570 jt above the dorsal wall of the sinus, as
a transversely elongated epithelial plate, and its cranial end reaches up for a short
distance as tapering 'wings' on either side of the lower end of the utero-vaginal
canal. Caudally it is continuous with the dorsal wall of the sinus, but its root is
split into two by a small mesodermal septum (PI. 2, fig. 9). Dorsal to this septum
the two roots join each other to form the sinus upgrowth, while on its ventral
aspect they are continuous with the enlarged dorsal and cranial portion of the sinus
which lies above and behind the urethro-vaginal folds. This arrangement obviously
cannot have arisen from conditions such as have been described in the 94 mm.
foetus. Apparently, as in the 68 mm. specimen, a small mesodermal septum must
have divided the root of the sinus upgrowth at the time of its initial formation.
The epithelium of the sinus upgrowth stains rather more deeply than in the
94 mm. foetus, but the basal and superficial zones can be readily recognized and it
can be distinguished from the Mullerian epithelium with which it is in contact (PI. 2,
fig. 10). At the caudal end of the upgrowth, and in the enlargement of the dorsal
portion of the sinus from which it arises, the epithelium is of the same differentiated
type, but here the pale-staining cells are much larger and more prominent than they
are further cranially, their cytoplasm eosinophilic and their nuclei very small (PI. 2,
fig. 9). The remainder of the pars pelvina is still lined by the same undifferentiated
type of epithelium as in the earlier stages.
496 D. Bulmer
The lower part of the utero-vaginal canal is lined by a stratified polygonal
epithelium, and the first sign of the differentiation of the surrounding mesoderm
which marks the position of the future external os (Bulmer, 1955) indicates that this
epithelium lines the entire 'vaginal' portion of the canal, the uterine segment
retaining its original columnar epithelium.
161 week foetus
The crown-rump length of this specimen is unfortunately not known, but it
shows a slightly later stage of development than the 112 mm. foetus. There is now
a marked dorso-ventral elongation of the pars phallica and a relative and absolute
shortening of the pars pelvina compared with the 94 mm. foetus. Though this
change of shape has slightly modified the relative positions of the folds in the sinus
wall their pattern remains essentially the same (Text-fig. 3).
UV

C d
a
a

m
fed C. e

ha

gX,..- as~~ ~ ~~~~~~~~ ,....

Text-fig. 3. 16j week foetus. Median sagittal section and the corresponding transverse sections
through the sinus and the lower end of the utero-vaginal canal ( x 13j). The arrow at s
indicates the level of the cranial tips of the sinus upgrowth, on either side of the lower end of
the utero-vaginal canal. Other lettering and shading as for Text-fig. 1.

The sinus upgrowth extends cranially for a distance of about 1*6 mm.-approxi-
mately half the extent of the future vagina as indicated by the site of the external os.
In most of its extent the upgrowth forms a solid transversely elongated epithelial
plate, and cranially it extends for a short distance as bilateral wings on either side
of the lower end of the utero-vaginal canal. Its epithelium is again differentiated
into basal and superficial zones, and the thickness of section (12 4a) seems to
accentuate the distinction between Mullerian and sinus epithelia (PI. 2, fig. 11).
The 'vaginal' portion of the utero-vaginal canal is lined by a four- or five-layered
stratified Mullerian epithelium, and the superficial cells are now becoming more
flattened.
At its caudal end the sinus upgrowth joins the enlargement of the sinus which
lies cranial and dorsal to the urethro-vaginal folds, unsplit by any mesodermal
septum, and this lower end of the upgrowth now shows three swellings of its epithelial
plate, one centrally and one on either side. In these swellings, and in the dorsal
 The development of the human vagina 497
enlargement of the sinus, the epithelium is rather different from that in the cranial
part of the sinus upgrowth. The internal cells are larger and the nuclei smaller, so
that the distinction between basal and superficial zones is much more apparent
(P1. 2, fig. 12). In addition, particularly in the swellings of the sinus upgrowth, the
cytoplasm of the internal cells is eosinophilic, and the cell walls deeply stained.
The remainder of the pars pelvina, below and ventral to the urethro-vaginal folds,
is still lined by an undifferentiated type of epithelium, consisting of four or five
layers of small, darkly staining cells.
140 mm. foetus
In this specimen many further developments are apparent. The sinus upgrowth
extends to the lower end of the cervical canal as a solid, transversely elongated
epithelial plate in which the basal and internal zones of cells can be distinguished.
Here it meets a stratified squamous Mullerian epithelium (P1. 2, fig. 13), which
consists of very much smaller cells and lines the lower portion of the cervical canal
immediately above the sinus upgrowth, intervening between this and the columnar
epithelium of the rest of the uterus above. The vaginal fornices have not yet
appeared.
As the sinus upgrowth is followed caudally, it develops, about the middle of its
cranio-caudal extent, enlargements similar to those seen at the lower end of the
upgrowth in the 16J week foetus-a large swelling centrally, and smaller swellings
at each lateral extremity. Traced further caudally these swellings become larger,
until eventually they merge together to form a single mass occupying about the
lower one-third of the vagina, rather heart-shaped in cross-section and with a large
central lumen (Text-fig. 4). In the swellings the vaginal epithelium consists of
a narrow basal zone of darkly staining cells, with relatively large nuclei, and a mass
of large internal pale-staining cells with very small nuclei, eosinophilic cytoplasm
and very deeply stained cell walls (P1. 3, fig. 14). In the canalized portion of the
vagina the epithelium is very similar, with a thick zone of internal cells of which the
superficial layers are flattened. The lumen is filled by a mass of desquamated
material, and the PAS technique demonstrates that all but the basal cells of the
epithelial lining are loaded with glycogen.
The lower end of the vagina communicates with the sinus by paired hymeneal
orifices (Text-fig. 4b and PI. 3, fig. 15), separated from each other by a small median
mesodermal septum, and the great enlargement of the lower end of the vagina has
resulted in the extension of its area of contact with the sinus, particularly on the
lateral and caudal aspects of the orifices (Text-fig. 4c). In this way the hymen is
formed, consisting of a plate of dense connective tissue, lined above by vaginal
epithelium and below by the undifferentiated type of sinus epithelium which still
occupies the upper part of the sinus (PI. 3, fig. 17).
Associated with the increasing area of contact between the vagina and the sinus
which gives rise to the hymen, the configuration of the sinus itself shows a consider-
able change from the stage represented by the 16j week foetus. The pars pelvina, as
judged by the site of entry of the ducts of Bartholin's glands (Text-fig. 4), is now
very short, and the original system of longitudinal folds, though it can still be
followed, has become modified by the change in shape of this part of the sinus.
498 D. Bulmer
A striking feature, which seems to be associated with the invagination of the upper
part of the dorsal wall of the sinus by the lower end of the vagina, is the accentuation
of the upper end of the lateral fold, now lying parallel with the posterior portion
of the hymen and separated from it by an upward extension of the dorso-lateral
sinus bay-the bay which, at a lower level (Text-fig. 4), receives the ducts of
Bartholin's glands. The dorso-lateral bay extends cranially and ventrally on the
lower aspect of the hymen as far as the hymeneal orifices, where it terminates as
the lateral fold becomes continuous with the hymeneal fold. In addition, it gives off
a cranial diverticulum which ascends for a short distance on the lateral aspect of

a ~~~~~~b c

md

hy ba

C h
e - ba \=

Text-fig. 4. 140 mm. foetus. Median sagittal section through the vagina and the sinus, with the
corresponding transverse sections (x 6j). The stippling indicates the vaginal epithelium,
and the vestibule is shown only in outline. Because of the dilatation of the upper end of the
sinus it is not practicable to indicate the folds of its wall in the median sagittal section.
However, the dorsal margin of the lower part of the lateral fold is shown by the dotted line,
and the small inset figure shows the outline (dotted line) of the upper end of the dorso-lateral
sinus bay and its cranial diverticulum in relation to the margins of the vaginal and sinus
epithelia (continuous lines) and to the hymeneal orifices (interrupted line). hy-hy' indicates
the level of the hymeneal orifices in relation to the median sagittal section. Other lettering
as for Text-fig. 1. It will be seen that the change in the position of the lateral fold in this
specimen, and to an even greater extent in the next (Text-fig. 5), differs from that described
by Mijsberg (1924). The dorso-lateral bay is rotated relatively upwards and backwards with
the downgrowth of the vagina into the sinus. According to Mijsberg, this bay is rotated
upwards and forwards, an opinion for which the present material offers no evidence.

the lower end of the vagina (PI. 3, figs. 15, 16), and presumably represents the
cranial extremity of the dorso-lateral bay of the 161 week stage, maintaining its
position despite the relative downgrowth of the lower end of the vagina.
The hymeneal folds, bounding the outer margins of the paired hymeneal orifices,
represent the dorsal margins of the urethro-vaginal folds of the earlier stages. In
the same way they form the boundary between the differentiated and undifferenti-
ated types of sinus epithelium, and it is by the enlargement of the area of dif-
ferentiated epithelium-now the lower end of the vagina-and its downgrowth
relative to the urogenital sinus that the hymen is formed. The mesodermal septum
between the two hymeneal orifices seems to differ in origin from the remainder of
 The development of the human vagina 499
the hymen, and probably results from the further development of a septum such as
occurs in the 112 mm. foetus, splitting the root of the sinus upgrowth. By the
enlargement and downgrowth of the lower end of the vagina the septum is pro-
jected forwards to divide the space between the dorsal margins of the urethro-
vaginal folds-in other words, to split the hymeneal orifice. The mesodermal septum
of the 140 mm. stage is therefore lined on both its ventral and dorsal aspects by
vaginal epithelium.
180 mm. foetus
The enlargement and canalisation of the vagina now extends throughout its
entire length, and the fornices are well established. An important change is the
disappearance of the zone of stratified squamous Mullerian epithelium from the
lower portion of the cervical canal, and the sinus epithelium of the vagina meets the
columnar epithelium of the uterus just inside the external os. Here there is a very
clear line of distinction between the two epithelia (PI. 3, fig. 18). Both give a strongly
positive PAS reaction, granular in the vaginal epithelium and diffuse in the columnar
cells of the cervix, but in the latter this reaction is not affected by previous salivary
digestion.
The vagina is lined by a very thick stratified squamous epithelium (PI. 3, fig. 19),
in which it is possible to distinguish four cellular zones. The basal zone is formed by
three or four layers of small cubical cells, with relatively large nuclei and basophil
cytoplasm. The next zone consists of seven or eight layers of much larger cells,
with relatively much smaller nuclei and pale, acidophil cytoplasm. The deeper
layers of this zone consist of polygonal cells, with deeply stained cell walls, but the
three or four superficial layers are formed by very flattened cells, with much
thicker cell walls. The third zone consists of about ten layers of large, clear-staining
polygonal cells, less flattened and with thinner cell walls than the most superficial
cells of the second zone. The nuclei are very small, and are absent from many of the
cells. The innermost zone is formed by four or five layers of cells, similar to those of
the third zone but more flattened. Many of the cells are without nuclei, and the
nuclei which are present are very small and pyknotic.
Glycogen is extremely abundant in the vaginal epithelium in all but the basal zone
of cells, as shown by staining with Best's ammoniacal carmine and by the PAS
technique. It is generally located as large granules, sometimes almost filling the
cytoplasm of the cell, and is usually, but not always, restricted to the part of the
cell on the proximal side of the nucleus. In addition, the cell walls of the three
superficial zones show a positive PAS reaction after previous salivary digestion,
similar to that demonstrated by Wislocki, Fawcett & Dempsey (1951) in the
vaginal epithelium of the adult.
A further development from the 140 mm. stage is the lining of the under surface
of the hymen and the adjacent portion of the vestibule by an epithelium similar to
the vaginal epithelium, though rather thinner. Associated with the further relative
downgrowth of the lower end of the vagina, and the resulting more horizontal
position of the hymen (Text-fig. 5), the pars pelvina has almost disappeared. In
fact, the shape of the sinus is so distorted from its original form that the distinction
between pars pelvina and pars phallica now has little meaning.
31 Anat. 91
500 D. Bulmer
By this stage the sinus extends further dorsally in relation to the under surface
of the hymen, and the dorso-lateral sinus bays are very prominent, forming a
dilatation of the vestibule immediately beneath the hymen (Text-figs. 5 a, b). They
extend further ventrally than in the 140 mm. foetus, reaching forwards around the
sides of the anterior portion of the hymen. As before, they send short diverticula
cranially, on either side of the lower end of the vagina. The hymenal orifice is
unsplit.
200 mm. foetus
This specimen shows little further change, but the sagittal plane of section gives
a very clear picture of the penetration of the lower end of the cervical canal by the
vaginal epithelium, to a distance of about 3 mm. above the external os.
375 mm. foetus
The vaginal epithelium is now considerably thinner, but the lumen is filled with
cell debris, suggesting that the decrease in thickness may be largely due to de-
squamation of the superficial cell layers. Only three cellular zones can be dis-
tinguished (P1. 3, fig. 20). In the basal zone there are usually three layers of small
cubical cells, with large nuclei and basophil cytoplasm. The intermediate zone is
U

a
b cd.
C

db

a b C
Text-fig. 5. 180 mm. foetus. Median sagittal section through the lower end of the vagina and the
vestibule, with the corresponding coronal sections ( x 4). The dotted line in the sagittal section
shows the relative extent of the dorso-lateral bays. The coronal sections are shown with their
cranial ends towards the top. u, urethra; v, vagina; h, hymeneal orifice; db, dorso-lateral
sinus bay; cd, cranial diverticulum of dorso-lateral sinus bay.

formed by four or five layers of much larger polygonal cells, with relatively smaller
nuclei and deeply staining cell walls. Superficially there are two or three layers of
large flattened cells with small pyknotic nuclei. The superficial and intermediate
zones are not clearly distinct from each other, and both are rich in glycogen granules.
It is possible that these two zones represent the second zone of the epithelium of the
180 mm. foetus, the third and fourth zones having been desquamated. Again the
cell walls give a positive PAS reaction after salivary digestion.
The sinus shows the further progress of the developmental changes noticed at the
180 mm. stage. The dorso-lateral bays extend still further forwards, now reaching
around the urethral orifice, and the dorsal extension of the sinus in relation to the
under surface of the hymen is still more marked. The whole of the upper part of the
 The development of the human vagina 501
sinus is lined by an epithelium similar to that of the vagina, and now of about the
same thickness.
1 month infant
Little change is noticed in this specimen, except in the character of the vaginal
epithelium (PI. 3, fig. 21). This is considerably thinner than in the later foetal stages
and consists mainly of polygonal cells with relatively large nuclei, deeply staining
cytoplasm and indistinct cell walls. Glycogen is no longer present, and the cell walls
no longer give a positive PAS reaction.

DISCUSSION
There can be little doubt that the epithelium of the human vagina is entirely
derived from an upgrowth of sinus cells, as Vilas (1932), Kempermann (1935) and
Meyer (1934-88) claimed, but the manner of formation of this upgrowth does not
seem to be completely in accord with their findings. Vilas believed that the epi-
thelium in the dorsal wall of the sinus proliferated in two different ways. An inner
proliferation of pale-staining cells grew ventrally and medially in the internal layers
of the dorsal sinus wall, meeting its fellow to form a crest which projected ventrally
into the sinus lumen. An outer proliferation of darkly staining cells grew dorsally,
displaced the Mullerian cells from the Mullerian tubercle and extended cranially as
an epithelial plate, subsequently canalized to form the vagina. The views of Meyer
and Kempermann were essentially similar.
It is interesting to compare this with Koff's account (1933) of the origin of the
'sino-vaginal bulbs', evaginations of the dorso-lateral aspect of the sinus on each
side. Koff found that their epithelium was different from that of the rest of the
sinus, consisting of well-marked basal and superficial zones, the basal cells darkly
staining and the superficial cells pale-staining. The bulbs became larger and
eventually fused with each other to form a solid cellular mass, which displaced the
Mullerian tissue dorsally and cranially and eventually became canalized as the
lower one-fifth of the vagina. The sino-vaginal bulbs seem to be the structures
which have here been termed the dorso-lateral projections. In the same way they
were joined by the lower ends of the Wolffian ducts, when these persisted, and
carried the attachments of the ducts with them as they extended cranially. Koff
did not mention any proliferation of darkly staining cells from the intermediate
portion of the dorsal wall of the sinus, such as occurs in the 65 mm., 68 mm. and
14-week foetuses of this present collection, and would presumably regard this
proliferation as an early stage in the fusion of the sino-vaginal bulbs.
Up to a point, the findings of this present investigation combine both these views
on the initial origin of the sinus contribution to the vagina, the sinus upgrowth
taking origin from all three of the elements associated with the activity of the
dorsal wall of the sinus, first noticed in the 65 mm. foetus. The junction of the
persistent lower end of the left Wolffian duct with the sinus upgrowth in the
94 mm. specimen indicates that, at any rate in this foetus, the dorso-lateral pro-
jection must have participated in the formation of the upgrowth. On the other
hand, in the 65 mm. foetus there seems no doubt as to the entity of the inter-
mediate proliferation of darkly staining cells, and their contribution to the sinus
31-2
502 Btulmer
D.
upgrowth cannot be ignored. Only by the fusion of the three initial elements, by
the 94 mm. stage, is the single sinus upgrowth produced. It is interesting that the
formation of the sinus upgrowth in the human foetus bears such a close similarity
to the initial origin of the lower vaginal segment in the sheep (Bulmer, 1956). There
may be no great significance in the distinction between the dorso-lateral projections
and the intermediate proliferation of darkly staining cells, all three representing a
continuous cellular proliferation from this area of the sinus wall. There seems little
doubt, however, of the identification of the dorso-lateral projections with Koff's
sino-vaginal bulbs, and of the intermediate proliferation with the outer sinus
proliferation of Vilas.
In many instances, it appears, the root of the sinus upgrowth is split at its origin
from the dorsal wall of the sinus by the inclusion of a small mesodermal septum.
Vilas pointed out that in his early foetuses, from the 38 mm. stage onwards, a small
mass of mesoderm interrupted the contact between the Mullerian epithelium and the
dorsal wall of the sinus in the midline, and this has been noticed here in the 28, 42
and 50 mm. foetuses. Vilas described the initial bilateral origin of the darkly staining
sinus proliferation, the two origins then fusing to form a single sinus upgrowth. In
the 65 mm. foetus, described here, there is no sign of such a bilateral origin, but in
the 68 mm. foetus the proliferation is split by a small mesodermal septum. The
inclusion of a mesodermal septum in the root of the sinus upgrowth is most prob-
ably a result of the persistence of such a septum at the Mullerian tubercle, and this
would suggest that the darkly staining proliferation only arises where the sinus is
actually in contact with Mullerian cells. The persistence of such a septum, as in the
112 mm. foetus, and its development into a hymeneal septum, as in the 140 mm.
foetus, must be fairly common, and occurred in many of the specimens examined by
Meyer (1934-38). The persistence of a more extensive septum in the utero-vaginal
canal would account, in a similar manner, for a congenital duplication of the vagina.
The further development of the sinus upgrowth, after the 94 mm. stage, is of some
interest. Vilas found that as the Mullerian epithelium receded before the advancing
plate of sinus epithelium the lower end of the utero-vaginal canal, including its
whole 'vaginal' portion, became completely occluded by the proliferation of its
lining cells. A solid vaginal plate was thereby formed, composed of Mullerian tissue
above and sinus tissue below, but the sinus epithelium gradually extended further
and further cranially until eventually the Mullerian epithelium was completely
displaced from the vaginal plate. In the series of foetuses examined here the
occlusion of the lower end of the utero-vaginal canal is not a prominent feature. It
is first seen at the 65 mm. stage and persists until the Mufllerian epithelium is
completely excluded from the vagina, but it never involves more than a short
segment. It is to be noted, however, that the stratification of the epithelium of the
utero-vaginal canal does extend, apparently by the 94 mm. stage, throughout the
entire 'vaginal' portion of the canal.
Koff also described a 'primitive vaginal plate', formed in its lower portion by the
solid epithelial mass of the sino-vaginal bulbs, and in its upper portion from the
occlusion of the 'vaginal' portion of the utero-vaginal canal by the proliferation of
its lining cells. In this solid vaginal plate Koff claimed that he was able to dis-
tinguish histologically between the sinus and Mullerian components, and only after
 The development of the human vagina 503
the 142 mm. stage was this distinction no longer apparent. He assumed that the
relative proportions of the two epithelia remained unchanged after this stage, the
sinus epithelium being restricted to the lower one-fifth of the vagina.
Unfortunately, Koff showed no microphotograph of the epithelial distinction
between the components of his primitive vaginal plate, but the criterion which he
used for the identification of the sinus epithelium was its differentiation into
external darkly staining and internal pale-staining zones. It has been noticed in
the series of foetuses described here that after the 94 mm. stage the internal cells in
the cranial part of the sinus upgrowth are much less pale-staining than they are
further caudally. For instance, in the 16j week foetus the large, markedly pale-
staining cells are restricted to the enlarged portion of the dorsal wall of the sinus
and to the swellings at the caudal end of the sinus upgrowth. Indeed, Koff seems
to have identified the two laterally placed swellings with the sino-vaginal bulbs
themselves-a conclusion which is difficult to follow and for which there is no
evidence in the material described here. The epithelium at the cranial end of the
sinus upgrowth is of the same essential character as that further caudally. There is
no evidence of any sharp histological distinction in the vaginal mass except the one
which has been identified as the junction of the cranial end of the sinus upgrowth
with the caudal end of the Mullerian tissue. Koff's description offers no expla-
nation for the very marked distinction between the two, occurring about half-way
up the 'vagina' in the 161 week foetus.
The most likely explanation for the histological differences between the cranial
and caudal ends of the sinus upgrowth is that the differentiation of its cells progresses
from below upwards. Thus, the cells of the caudal end of the upgrowth in the 112 mm.
and 161 week foetuses, when the internal cells are larger, their nuclei smaller and the
cell walls deeply stained with eosin, are similar to those occupying the whole of the
lower half of the vagina in the 140 mm. foetus. The transversely elongated plate of
the vagina appears to differentiate, at each particular stage, by first enlarging to
form swellings. The epithelium correspondingly proliferates and the internal cells
become large, with small nuclei and deeply stained cell walls. The swellings become
still larger, and confluent with each other, the central cells then desquamating to
form the vaginal lumen. The 140 mm. foetus shows later and later stages of this
process as the sinus upgrowth is followed further and further caudally. The vaginal
epithelium continues to proliferate, presumably under the influence of hormonal
stimulation, and it is interesting to compare its structure in the two older foetuses
with the adult vaginal epithelium as described by Papanicolou, Traut & Marchetti
(1948), and more particularly, in view of the probable hormonal influences at work,
with the vaginal epithelium of the pregnant woman, described by Smith & Brunner
(1934).
As indicated above, many earlier workers have denied the participation of sinus
epithelium in the formation of the vagina. Berry Hart (1896, 1901 and 1911) and
Mijsberg (1924) described a contribution to the vagina from the 'Wolffian bulbs',
proliferations of the epithelium of the lower ends of the Wolffian ducts. There
seems no doubt that the Wolffian bulbs were the dorso-lateral projections, and
there is no evidence that these latter arise from Wolffian epithelium. Their cells are
unlike the cubical Wolffian cells which lie dorsal to them, and much more closely
504 D. Bulmer
resemble those of the dorsal sinus wall. Berry Hart believed that the epithelium
of the Wolffian bulbs extended throughout the entire vagina, while Mijsberg, in
a sense less correctly, restricted the Wolffian contribution to a lower vaginal
segment. His reasons for this were similar to those which influenced Koff to believe
that the sino-vaginal bulbs formed only the lower end of the vagina.
Other authors have held to the view that the human vagina is entirely derived
from the Mullerian utero-vaginal canal. Bloomfield & Frazer (1927) illustrated a
section rather similar to the ones shown in PI. 1, figs. 4, 5 and 6, for the 65 mm.
foetus, but their interpretation was very different, as they believed that the pale-
staining cells in the dorsal sinus wall were the Mullerian cells breaking through into
the sinus at the Mullerian tubercle. It is clear that the findings in the 65 mm. foetus
make such a view untenable, and this misinterpretation presumably led Bloomfield &
Frazer to think that the whole extent of the differentiated type of sinus epithelium
was of Mullerian origin.
While the sinus origin of the vaginal epithelium appears to be established, one
or two points of detail merit further consideration. From the earliest stage
examined, represented by the 28 mm. foetus, the epithelium which lines the part
of the sinus from which the sinus upgrowth later arises differs from that of the rest
of the pars pelvina in consisting of two cellular zones-darkly staining cells deeply
and pale-staining cells superficially. So far, it has been tacitly assumed that both
these cell types are of sinus origin, but other workers have placed different inter-
pretations upon them. Vilas (1932) believed that the darkly staining cells were of
sinus origin, but was prepared to accept that the pale-staining cells were Wolffian,
growing ventrally and medially in the dorsal wall of the sinus to form the ventrally
projecting Wolffsche Kamm, identified by Kempermann (1931) and evident in the
65 mm. foetus of this present investigation. Kempermann (1935) also believed that
the pale-staining cells were Wolffian in origin, but that they later completely
disappeared. The pale-staining cells in the internal layers of the sinus upgrowth at
a stage such as that represented by the 94 mm. foetus of this collection were not the
Wolffian cells, but a new generation of cells derived from the sinus. Meyer (1934-38),
however, believed that the pale-staining Wolffian cells did persist into later foetal
life, identifying them with the large pale-staining cells such as occur at the root of
the sinus upgrowth in the 112 mm. and 161 week foetuses.
This question does not seem possible of solution by ordinary histological methods.
Nevertheless, the pale-staining cells in the dorsal wall of the sinus from the 28 mm.
stage onwards bear no obvious similarity to the cubical cells which line the Wolffian
ducts, and are distinct from the Wolffian cells which lie dorsal to them. It appears
just as reasonable to suppose that the pale-staining cells are derivatives of the
darkly staining basal cells, and that there is no incorporation of Wolffian cells in
the dorsal wall of the sinus at this stage. No such ingrowth occurs in other mammals,
and it seems unnecessary to postulate it in the human foetus. Nevertheless, it must
be admitted that the differentiation of the sinus epithelium from which the sinus
upgrowth arises is a remarkable feature, and it too has not been described in other
mammals.
Also of some significance is the problem of the behaviour of the lower ends of the
Wolffian ducts during the development of the vagina. Descriptions of the lower ends
 The development of the human vagina 505
of the Wolffian ducts in female foetuses fall generally into two main categories. In
the first are those which maintain that the Wolffian ducts continue to open into the
sinus at their original site, immediately lateral to the position of the Mlilerian
tubercle. Thus, Meyer (1909) described the persistence of the lower ends of the
Wolffian ducts in a large number of older foetuses and new-born children, running
into the substance of the hymen and opening into the vestibule close beside the
hymeneal orifice. On the other hand, many other workers (Tourneux & Legay, 1884;
van Ackeren, 1889; Mijsberg, 1924; Koff, 1933) have described the cranial migration
of the lower ends of the Wolffian ducts to join the lower end of the vagina itself,
though the ducts degenerated very soon after and did not persist into late foetal life.
At first sight, the present investigation supports this latter view. In the 94 mm.
foetus the lower end of the left Wolffian duct joins the sinus upgrowth, indicating
that the dorso-lateral projection has carried the Wolffian remnant with it as it grew
cranially, while in none of the older foetuses was there any trace of the lower ends
of the Wolffian ducts. Nevertheless, this evidence is insufficient to suggest that the
lower ends of the Wolffian ducts never maintain their openings into the sinus beside
the original site of the Mullerian tubercle, particularly in view of the large number of
such cases described by Meyer (1909). At that time Meyer regarded these Wolffian
remnants as indicative of the entirely Mullerian origin of the vagina, but in his later
publications (1934-38) pointed out that they were equally in accord with the account
of vaginal development given by Vilas. The Wolffian ducts were not involved in the
proliferation which Vilas believed to form the epithelial plate of the vagina, and
were therefore left behind near the hymen. Nevertheless, the structures which
Meyer now identified as persistent Wolffian remnants do not appear to occupy the
same site as those which he had described earlier, but join the lower end of the
vagina on the upper aspect of the hymen. In other words, they have been carried
cranially for a short distance during the formation of the sinus upgrowth.
It may be that in some instances the lower ends of the Wolffian ducts are not
carried cranially with the sinus upgrowth to such a great extent as they have been
in the 94 mm. foetus, or indeed that they are not carried cranially at all. This may
be because the dorso-lateral projections do not play such a large part in the forma-
tion of the sinus upgrowth in these cases, or because the manner of their growth is
such as not to involve the lower ends of the Wolffian ducts. In other words, it is
possible, as von Lippmann suggested (1939), though with rather a different purpose,
that there is some variability in the behaviour of the lower ends of the Wolffian
ducts during the early stages of vaginal development. This is supported by the fact
that an ectopic ureter, presumably opening by a persistent Wolffian remnant, may
join either the vagina or the vestibule (Kermauner, 1909). It is also possible that
many of the structures which have been identified as persistent Wolffian remnants
were, in fact, not so. In the 140 mm. foetus the dorso-lateral sinus bays end above
by sending short diverticula upwards on either side of the lower end of the vagina,
lined by the undifferentiated type of sinus epithelium. In the 180 mm. foetus these
diverticula are still lined by the undifferentiated epithelium, though the dorso-
lateral bays themselves are occupied by an epithelium very similar to that of the
vagina. The diverticula are presumably associated with the relative downgrowth of
the lower end of the vagina, as has already been pointed out. They are, however,
506 D. Bulmer
very similar to the 'WoIffian remnant' described by Bloomfield & Frazer (1927), in
a 170 mm. foetus of their collection, but their lining epithelium leaves no doubt of
their origin (P1. 3, fig. 16).
An interesting viewpoint on the development of the vagina was suggested by
Zuckerman (1940), who believed that any epithelium of the adult genital tract which
responded to oestrogenic stimulation by a stratified squamous proliferation-a
'squamous response'-was a derivative of the sinus epithelium of the foetus. The
conclusion of this present investigation, that the entire vaginal epithelium is derived
from the sinus upgrowth, might appear to support Zuckerman's hypothesis. There
is, however, a considerable body of embryological evidence which indicates that in
many other mammalian forms the upper vaginal segment, though it is lined by a
stratified squamous epithelium and gives a squamous response to oestrogenic
stimulation, is a Mullerian derivative. Thus, it seems likely that when the Mullerian
epithelium does persist in the vagina it tends to be of stratified squamous form and
its response to oestrogenic stimulation squamous in type. Of some interest in this
respect is the lining of the 'vaginal' portion of the utero-vaginal canal in the human
foetus by a stratified squamous epithelium of Mullerian origin, and the disappearance
of this epithelium after the 140 mm. stage, when it is displaced into the cervical
canal. In the absence of any firm knowledge of the hormonal conditions in female
human foetuses, we do not know how this stratified Mullerian epithelium responds
to oestrogenic stimulation. Probably, however, there is some controlling factor,
producing its effect in foetal life in the human subject and throughout life in many
other mammals, which stimulates Mullerian cells in the vagina, but not in the
uterus, to form a stratified squamous epithelium.
One of the most interesting problems is that of the extent to which the develop-
ment of the human vagina is influenced by hormonal factors. It is a reasonable
supposition that the enormous activity of the vaginal epithelium in the older
foetuses is a result of stimulation by maternal oestrogens, as Fraenkel & Papanicolou
(1938) suggested. The gross differences between the vaginal epithelium in these
foetuses and that of the 1-month old infant imply that this oestrogenic stimulation
must extend over a considerable period of foetal life-at any rate from the 112 mm.
stage onwards. The question of the hormonal control of genital development has
introduced a new field of experimental embryology, in which a very large amount
of work has already been carried out (for bibliography see Jost, 1948), and it raises
many interesting problems concerning the development of the human vagina. It
may be that the enormous enlargement of the lower end of the vaginal mass at the
140 mm. stage, associated with the formation of the hymen, is a result of hormonal
stimulation to which only the differentiated type of sinus epithelium is sensitive.
At an earlier stage, the extension of the sinus upgrowth itself may be due to a
similar selective response of the differentiated epithelium. Some of the marked
differences between vaginal development in man and in other mammals may result
from differences in the degree of hormonal stimulation at various critical periods of
foetal life, rather than from inherent differences in the structures which go to form
the vagina. On the other hand, the histological differentiation of the epithelium of
the sinus upgrowth in the human foetus may be associated with an increased
sensitivity to hormonal stimulation, compared with other mammalian forms. While
 The development of the human vagina 507
such suggestions can be only tentative, it is, nevertheless, very likely that the
endocrinology of vaginal development holds the key to many of these outstanding
problems.
SUMMARY
1. The development of the human vagina has been studied from a series of female
foetuses ranging from 28 to 375 mm. crown-rump length, and from a 1-month old
infant.
2. An upgrowth arises from an area of 'differentiated' epithelium in the dorsal
wall of the sinus, in a manner which might be said to combine the descriptions of
Vilas (1932) and Koff (1933).
3. The sinus upgrowth extends throughout the entire region of the vagina by the
140 mm. stage, and forms the whole of its epithelial lining. The changes in the
vaginal epithelium of older foetuses are also described.
4. The findings are discussed in relation to the question of an early ingrowth of
Wolffian epithelium into the dorsal wall of the sinus, the problem of the behaviour
of the lower ends of the Wolffian ducts in female foetuses, and recent opinions on
the significance of the sex hormones in genital development.
My thanks are due to Prof. J. D. Boyd, for the loan of most of the material which
I was able to study. I am also indebted to Messrs J. F. Crane and A. Cain for the
microphotographs, and to Mr W. Cruickshank for his assistance with the diagrams.

REFERENCES
ACKEREN, F. VAN (1889). Beitrage zur Entwicklungsgeschichte der weiblichen Sexualorgane des
Menschen. Z. wiss. Zool. 48, 1-46.
BANKS, W. M. (1864). On the Wolffian Bodies of the Foetus. Edinburgh.
BLOOMFIELD, A. & FRAZER, J. E. (1927). The development of the lower end of the vagina. J. Anat.,
Lond., 62, 9-32.
BOLK, L. (1907). Beitrage zur Affenanatomie VI. Zur Entwicklung und vergleichenden Anatomie
des Tractus urethro-vaginalis der Primaten. Z. Morph. Anthr. 10, 250-316.
BULMER, D. (1955). The Development of the Vagina. Thesis for the degree of M.D., University of
Cambridge.
BULMER, D. (1956). The early stages of vaginal development in the sheep. T. Anat., Lond., 90,
123-134.
FELIX W. (1912). The development of the urogenital organs. In Keibel and Mall, Manual of
Human Embryology, Vol. II, p. 752. Philadelphia and London.
FRAENKEL, L. & PAPANICOLOu, G. N. (1938). Growth, desquamation and involution of the
vaginal epithelium of human foetuses and children, with a consideration of the related
hormonal factors. Amer. J. Anat. 62, 427-451.
GoMORI, G. (1952). Microscopic Histochemistry. University of Chicago Press.
HART, D. B. (1896). Preliminary note on the development of the clitoris, vagina and hymen.
J. Anat., Lond., 31, 18-28.
HART, D. B. (1901). Morphology of the human urogenital tract. J. Anat., Lond., 35, 330-375.
HART, D. B. (1911). Adenoma vaginae diffusum (adenomatosis vaginae), with a critical discussion
of present views of vaginal and hymeneal development. Edinb. med. J. 6, 577-590.
HUNTER, R. H. (1930). Observations on the development of the human female genital tract.
Contr. Embryol. Carneg. Inst. 22, 91-108.
JOST, A. (1948). Le contr6le hormonal de la differenciation du sexe. Biol. Rev. 23, 201-236.
KEMPERMANN, C. T. (1931). Beitrag zur Frage der Genese der menschlichen vagina. Gegenbaurs
Jb. 66, 485-531.
508 D. Bulmer
KEMPERMANN, C. T. (1935). Beitrage zur Entwicklung des Genitaltraktus der Sauger. III. Das
Schicksal der kaudalen Enden der Wolffsche Gange beim Weibe und ihre Bedeutung fur die
Genese der Vagina. Gegenbaurs Jb. 75, 151-179.
KERMAUNER, F. (1909). Die Missbildungen der weiblichen Geschlechts-organe. In Schwalbe,
Die Morphologie der Missbildungen des Menschen und der Tiere, Jena. T. III, p. 258.
KOFF, A. (1933). Development of the vagina in the human foetus. Contr. Embryol. Carneg. Instn,
24, 59-91.
LIPPMANN, R. VON (1939). Beitrag zur Entwicklungsgeschichte der menschlichen Vagina und des
Hymen. Z. Anat. EntwGesch. 110, 264-300.
MEYER, R. (1909). Zur Kentniss des Gartnerschen (oder Wolffschen) Ganges besonders in der
Vagina und dem Hymen des Menschen. Arch. mikr. Anat. 73, 751-792.
MEYER, R. (1934-1938). Zur Frage der Entwicklung der menschlichen Vagina. I. Arch. Gynaek.
158, 639-738; II, Arch. Gynaek. 163, 205-308; III, Arch. Gynaek. 164, 207-357; IV, Arch.
Gynaek. 165, 504-590; V, Arch. Gynaek. 167, 306-338.
MIJSBERG, W. A. (1924). tber die Entwicklung der Vagina, des Hymen und des Sinus urogenitalis
beim Menschen. Z. Anat. EntwGesch. 74, 684-760.
MULLER, J. (1830). Bildungsgeschichte der Genitalien. Dusseldorf: Arnz.
PAPANICOLOU, G. N., TRAUT, H. F. & MARCHETTI, A. A. (1948). The Epithelia of Woman's
Reproductive Organs. Commonwealth Fund, New York.
POLITZER, G. (1952). Das Schicksal des Sinus urogenitalis beim Weibe. Z. mikr-anat. Forsch. 59,
6-28.
POLITZER, G. (1955). Zur normalen und abnormen Entwicklung der menschlichen Scheide. Anat.
Anz. 102, 271-278.
RATHKE, H. (1832). Abhandlungen zur Bildungs- und Entwicklungsgeschichte des Menschen und der
Thiere. Leipzig.
RETTERER, E. (1891). Sur l'origine du vagin de la femme. C.R. Soc. Biol., Paris, Ser. 9,3,291-293.
SMITH, B. G. & BRUNNER, E. K. (1934). The structure of the human vaginal mucosa in relation to
the menstrual cycle and to pregnancy. Amer. J. Anat. 54, 27-85.
SPULER, A. (1930). Entwicklungsgeschichte des weiblichen Genitalapparates. In Veit-Stoeckel,
Handbuch der Gyndikologie, 3rd ed., Band I, p. 470 (quoted by Vilas). Wiesbaden.
TOURNEUX, F. & LEGAY, C. (1884). Memoire sur le developpement de l'uterus et du vagin. J. Anat.
Paris, 20, 330-386.
VALENTIN, G. G. (1835). Handbuch der Entwicklungsgeschichte des Menschen. Berlin.
ViLAs, E. (1932). tber die Entwicklung der menschlichen Scheide. Z. Anat. EntwGesch. 98,
263-292.
WISLOCKI, G. B., FAWCETT, D. W. & DEMPSEY, E. W. (1951). Staining of stratified squamous
epithelium of mucous membranes of Man and Monkey by the periodic acid-Schiff method.
Anat. Rec. 110, 359-376.
ZUCKERMAN, S. (1940). The histogenesis of tissues sensitive to oestrogens. Biol. Rev. 15, 231-271.

EXPLANATION OF PLATES
PLATE 1
Fig. 1. 28 mm. embryo. Transverse section through the openings of the Wolffian ducts into the
urogenital sinus. The sinus epithelium consists of basal deeply staining and superficial pale-
staining cells, and the pale-staining cells extend for a short distance into the lower ends of the
Wolffian ducts. The Mullerian ducts are in close relation with the Wolffian ducts, but a small
mass of mesoderm separates them from the sinus. ( x 215.)
Fig. 2. 48 mm. foetus. Sagittal section through the opening of the right Wolffian duct. The
differentiation of the sinus epithelium in this region can be seen, and the extension of the
pale-staining cells into the lower end of the Wolffian duct. ( x 215.)
Fig. 3. 65 mm. foetus. Transverse section through the genital cord. The Wolffian ducts are very
small structures, on either side of the utero-vaginal canal. ( x 215.)
Fig. 4. 65 mm. foetus. Transverse section through the junction of Mullerian and sinus epithelia.
A mass of darkly staining cells (d), arising from the dorsal wall of the sinus, separates the
caudal end of the Mullerian epithelium (m) from the sinus lumen. On the right side the section
passes below the bulk of the dorso-lateral projection, and the Wolffian duct is large. On the
left the Wolffian duct is a small structure, applied to the dorsal aspect of the dorso-lateral
projection (dp). ( x 215.)
Journal of Anatomy, Vol. 91, Part 4 Plate I

.MW

BULMER TIlE DEVELOPMENT OF TIHE HUMAN VAGINA

(Facing p. 508)
 Journal of Anatomy, Vol. 91, Part 4 Plate 2

-N .Y
.*411w
11 .., l,
_W *
,#Of 1r 3W:%
- , 0p;8t.1
p'., MO
4%4
'}
10' W#
iw

I
L. ~
vv*s..
~ ~ ~ ~
~~ ~ ~ ~ ~ ~ ~ ~ ~ A

W4r'*vt*;
.,

4 'r
04.0 -.

1.

9 n
9,S+ ,*v

. . -

n'412
m

BULAIER-THE DEVELOPMENT OF THE HUMAN VAGINA

 Journal of Anatomy, Vol. 91, Part 4 Plate 3

9042
_.

, 4i

..f

AJ
I., 4.
*I .
1
.I,

J
,.

is
19

v-

W v
6- -W.
O.:'.*

.1..145
4,* 1A
t. ...

0
.. fi '.
0.R ;' iF' i!t hAr
X,::
A n;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1 21

BULMER-THE DEVELOPMENT OF THE HUMAN VAGINA

 The development of the human vagina 509
Fig. 5. 65 mm. foetus. Transverse section 16 g4 caudal to Fig. 4. The Mfillerian epithelium now
forms a smaller mass dorsally, and a few of the darkly staining sinus cells separate it from the
pale-staining sinus epithelium ventrally. ( x 215.)
Fig. 6. 65 mm. foetus. Transverse section 8 /z caudal to Fig. 5. ( x 215.)
PLATE 2
Fig. 7. 94 mm. foetus. Coronal section through the cranial and dorsal enlargement of the sinus
which forms the root of the sinus upgrowth. The epithelial differentiation is very well marked.
(x84.)
Fig. 8. 94 mm. foetus. Coronal section through the sinus upgrowth, 90 1z dorsal to Fig. 7. The
small darkly staining cells in the centre of the section (m) represent the caudal tip of the
Mullerian epithelium. A thin layer of darkly staining cells forms the outer layer of the sinus
upgrowth, and there is a small mass of these cells on the right (to the left of the photograph).
90 ,t further dorsally, a similar small mass, on the other side, is joined by the persistent lower
end of the left Wolffian duct. ( x 215.)
Fig. 9. 112 mm. foetus. Transverse section through the dorsal part of the sinus, showing the two
roots of the sinus upgrowth separated from each other by the mesodermal septum (ins).
The pale-staining internal cells of the sinus are very prominent. ( x 215.)
Fig. 10. 112 mm. foetus. Transverse section through the vaginal mass, 288 /Z cranial to Fig. 9.
The sinus upgrowth forms two lateral wings, while the central portion (m) is formed by the
paler-staining polygonal cells of the caudal end of the Mfillerian tissue. ( x 215.)
Fig. 11. 164 week foetus. Transverse section through the junction of the Muillerian epithelium of
the utero-vaginal canal (m) with the cranial wing of the sinus upgrowth on the left side.
( x 215.)
Fig. 12. 164 week foetus. Transverse section through the sinus showing the junction of its dorsal
wall with the caudal end of the sinus upgrowth. The central enlargement of the upgrowth is
immediately dorsal to the sinus lumen, and a further enlargement projects dorsally from it.
The large clear-staining internal cells are prominent in the central enlargement. In the
dorsal enlargement the cytoplasm and cell boundaries of the internal cells are markedly
eosinophilic. ( x 84.)
Fig. 13. 140 mm. foetus. Transverse section through the left side of the cervical canal, showing the
junction between the stratified Mullerian epithelium (m) and the cranial tip of the sinus
upgrowth. ( x 160.)
PLATE 3
Fig. 14. 140 mm. foetus. Transverse section through the left vaginal enlargement, just below the
middle of the cranio-caudal extent of the vagina. The internal cells are large, their nuclei
relatively small and the cell-boundaries markedly eosinophilic. ( x 215.)
Fig. 15. 140 mm. foetus. Transverse section through the urogenital sinus and the lower end of the
vagina, at the level of the paired hymeneal orifices. On each side of the ventro-lateral aspect
of the vagina is the small cranial extremity of the dorso-lateral sinus bay (db). ( x 30.)
Fig. 16. 140 mm. foetus. Transverse section through the wall of the dorso-lateral sinus bay,
showing its lining of undifferentiated sinus epithelium. ( x 450.)
Fig. 17. 140 mm. foetus. Transverse section through the lower portion of the hymen, showing the
vaginal epithelium above, separated by the hymeneal connective tissue from the undifferenti-
ated sinus epithelium below. ( x 450.)
Fig. 18. 180 mm. foetus. Transverse section through the junction of the vaginal epithelium with
the columnar epithelium of the cervix. ( x 120.)
Fig. 19. 180 mm. foetus. Section through the vaginal epithelium. ( x 280.)
Fig. 20. 375 mm. foetus. Section of the vaginal epithelium overlying a mesodermal papilla.
( x 280.)
Fig. 21. Vaginal epithelium in the 1-month old-infant. The epithelium is not of uniform thickness,
and this photograph is taken from an area where it is rather thicker than usual, in order to
give a better impression of its histological structure. ( x 320.)