Revision as of 13:35, 8 September 2015

Embryology - 5 May 2024 Expand to Translate
Google Translate - select your language from the list shown below (this will open a new external page)
العربية \| català \| 中文 \| 中國傳統的 \| français \| Deutsche \| עִברִית \| हिंदी \| bahasa Indonesia \| italiano \| 日本語 \| 한국어 \| မြန်မာ \| Pilipino \| Polskie \| português \| ਪੰਜਾਬੀ ਦੇ \| Română \| русский \| Español \| Swahili \| Svensk \| ไทย \| Türkçe \| اردو \| ייִדיש \| Tiếng Việt These external translations are automated and may not be accurate. (More? About Translations)

Papanicolaou GN. The Sexual Cycle in the Human Female as revealed by Vaginal Smears. Am J Anat. 1933;52: 519–637.

Historic Embryology Papers

Historic Disclaimer - information about historic embryology pages
Pages where the terms "Historic" (textbooks, papers, people, recommendations) appear on this site, and sections within pages where this disclaimer appears, indicate that the content and scientific understanding are specific to the time of publication. This means that while some scientific descriptions are still accurate, the terminology and interpretation of the developmental mechanisms reflect the understanding at the time of original publication and those of the preceding periods, these terms, interpretations and recommendations may not reflect our current scientific understanding. (More? Embryology History \| Historic Embryology Papers)

Online Editor

These are currently some of the plates from the historic paper by George Papanicolaou (1883 – 1962) the pathologist who developed the diagnostic "Pap smear" test named after him. Originally a Greek clinician (Georgios Nikolaou Papanikolaou) he travelled throughout Europe before becoming in 1913 an American researcher. His first studies were in other species (Guinea-Pig) before extending the studies onto humans. This paper described the appearance of normal cell smears during the different stages of the menstrual cycle. The true value of the "Pap smear" test was the identification of abnormal cells associated with cytopathology of cancer of the cervix.

The original black and white photomicrographs were hand-coloured by the author to match his microscopic observations.

Historic Embryology - Menstrual
1839 Corpus Luteum Structure \| 1851 Corpus Luteum \| 1933 Pap Smear \| 1937 Corpus Luteum Hormone \| 1942 Human Reproduction Hormones \| 1951 Corpus Luteum \| 1969 Ultrastructure of Development and Regression \| 1969 Ultrastructure during Pregnancy

The Sexual Cycle In The Human Female As Revealed By Vaginal Smears

George N. Papanicolaou

Department of Anatomy, Cornell University Medical College, and Woman’s Hospital, New York City

This work has been aided by the Committee for Research on Sex Problems of the National Research Council, and by the National Committee on Maternal Health.

Three Figures and Ten Plates (Eighty-One Figures)

I. Introduction

The study of the female sexual functions in mammals has been greatly stimulated and advanced in recent years by the application of the vaginal smear method. This method, as originally applied to the guinea pig by Stockard and Papanicolaou in 1917, consists in the microscopic examination of smears prepared at frequent intervals from the ﬂuid content of the vagina. The vaginal ﬂuid usually has a mucous consistency and contains a variety of desquamated cells, as well as leucocytes, lymphocytes, often erythrocytes, and a large number of bacteria. As the relative number and the distribution of these elements change periodically, smears prepared from such ﬂuid show modiﬁcations in their composition and structure. The successive alternation of periods of sexual activity and inactivity, which characterizes the mammals, imparts to the vaginal ﬂuid a rhythmical sequence of typical cellular stages which can be easily recognized.

These cyclic changes affect the entire genital tract, and, consequently, every change in the vaginal ﬂuid is strictly correlated With corresponding changes in the other organs of the female genital system, particularly the uterus and the ovaries. The time of ovulation may be accurately detected by this method in living mammals, while, before its application, no such information could be obtained without an operation or the sacriﬁce of an animal.

In the guinea pig, which is polyoestrous, the sexual cycles return periodically throughout the year every 15 to 16 days (15.73 in average). The longest part of each sexual cycle, i.e., about 12 days, is occupied by a period of relative inactivity or rest, which is called ‘clioestrus.’ During this time the vaginal smear, as described by Stockard and Papanicolaou ( ’l7), consists chieﬂy of leucocytes and a varying num- ber of atypical squamous cells. The period of increased sexual activity lasts 3 to 4 days and is characterized by a succession of stages, which have been designated as stages I, II, III, IV. During stage I the leucocytes almost disappear from the smear, the secretion of mucus becomes more abundant, and the cells which dominate the smear are of a squamous type with very small pyknotic nuclei which are at times fragmented. At the end of stage I there is an intermediate period, characterized by the prevalence of ‘elongate, corniﬁed cells Without nuclei.’ Stage II also shows a scarcity of leucocytes and a prevalence of cells which are derived from the deeper layers of the vaginal epithelial wall, being thus less de- generated and having a larger nucleus and a more compact form. Stage III is characterized by the reappearance of myriads of leucocytes and by cells mainly of the II and III type. The III cells are as a rule modiﬁed type II cells, the bodies of which had been penetrated by the invading leucocytes. Stage IV is practically the same as stage III, with the difference that erythrocytes are also present in the smear as the result of slight bleedings.

These four smear stages are of short duration, succeeding one another rapidly and are strictly coordinated with corresponding stages of the uterine and ovarian cycles. During stage I the ovary contains large ripening follicles and regressing corpora lutea. The uterus is congested and hypertrophied and its epithelial lining consists of high cuboidal or columnar cells. During the intermediate or corniﬁed stage the follicles as well as the uterus reach their highest development. The onset of the catabolic processes, i.e., the bursting of the ripe follicles and the denudation of the uterine mucosa, is associated with the appearance of vaginal stages II and III. Ovulation in the guinea pig is thus characterized by deﬁnite vaginal changes and occurs at about the time of the sloughing off of the uterine mucosa. Comparable conditions have been found to exist in other rodents, such as the rat and the mouse.

Long and Evans (’22), in their monograph on the oestrous cycle in the rat, described deﬁnite relations between the vaginal cycle, as revealed by vaginal smears, and the uterine and ovarian cycles. The rat has normally a very short sexual periodicity of about 5 days’ duration. Long and Evans recognized ﬁve smear stages: Stage 1: Leucocytes disappear; great numbers of small round nucleated epithelial cells of strikingly uniform appearance and size; duration 12 hours. Stage 2, or ‘corniﬁed cell stage’: leucocytes still absent; appearance of corniﬁed cells. Stage 3, or ‘late corniﬁed cell stage’: leucocytes absent; rich in corniﬁed cells, forming large cheesy masses; duration of stages 2 and 3 about 30 hours. Stage 4, or ‘leucocytic—corniﬁed cell stage’: reappearance of leucocytes; gradual disappearance of the corniﬁed cells and appearance of epithelial cells; duration 6 hours. Stage 5 corresponds to the dioestrus, which is of an approximate duration of 3 days and is characterized by the presence of leucocytes and atypical epithelial cells in the vaginal smear.

The corniﬁcation of the vagina is much more pronounced in the rat than in the guinea pig and this accounts for some of the smear differences between these two animals. In the rat the corniﬁcation extends over two well-deﬁned stages, the second and third of Long and Evans, whereas in the guinea pig it is usually overlapped by the I and II to the III stages; on this account, it had been originally described by Stockard and Papanicolaou as an intermediate period. In the rat the uterine growth is completed during the ﬁrst and second stages. The catabolic processes in the uterus begin during the corniﬁed cell stages, and the ovulation occurs during the last hours of this same period. The uterine epithelium is vacuolized and degenerated, but an actual denudation, as in the guinea pig, does not occur. The various phases of the sexual cycle in the rat proceed in rapid succession, but their coordination is maintained.

The corresponding vaginal smear stages in the guinea pig and the rat could be illustrated as follows:

Vaginal Smear Comparison Table
Guinea pig	Rat (Long and Evans)
I. Superﬁcial squamous cells with pyknotic nuclei; progressive leucopenia.	1. Small round nucleated cells; disappearance of leucocytes.
Intermediate (cornification) period a. Many corniﬁed cells, usually mixed with type I cells; b. Prevalence of corniﬁed cells with cheesy masses; leucopenia	Cornification period 2. Early corniﬁed cell stage; leucocytes scarce. 3. Late corniﬁed cell stage with large cheesy masses; leucopenia
II, III, IV. Appearance of deep layer cells; reappearance and great exodus of leucocytes; gradual disappearance of corniﬁecl cells; sometimes erythrocytes present.	4. Leucocytic-corniﬁed cell stage; reappearance of leucocytes; gradual disappearance of the corniﬁed cells.
V. Dioestrus: Leucocytes and atypical atypical vaginal cells.	5. Dioestrus: Leucocytes and vaginal cclls.
Reference: Papanicolaou GN. The Sexual Cycle in the Human Female as revealed by Vaginal Smears. Am J Anat. 1933;52: 519–637. Links: Estrous Cycle \| Guinea Pig Development \| Rat Development

This comparison indicates that the succession of stages is similar in both animals. Ovulation occurs at practically the same moment in relation to the above stages, i.e., at the end of the intermediate period in the guinea pig and during the last hours of the corniﬁcation period (second and third) in the rat.

In the mouse, Allen (’22) has also recognized ﬁve stages, the entire sexual cycle averaging about 4% days.

Selle (’22) has reclassiﬁed the vaginal smears of the guinea pig after a detailed study of the changes in the vaginal epithelium. He has recognized a separate ‘corniﬁed cell stage’ similar to the same stage in the rat. Such a stage may be actually present in a large number ‘of cases, but, as a rule, it is indistinct and overlaps both the preceding and the following stages. This corniﬁed cell stage of Selle corresponds to the ‘intermediate period’ of Stockard and Papanicolaou, and not to the stage 2 as given in Selle’s table. His stage 3 corresponds to the intermediate +II, whereas his stage 4 is equivalent to the stages III and IV of Stockard and Papanicolaou.

Murphey (’22) and Frei and Metzger (’26) have recognized four stages in the oestrous cycle in the cow with the vaginal smear method. Hartman (’23), in his study of the oestrous cycle in the opossum, also recognized four stages corresponding to the ones described in the rodents.

In 1924, McKenzie and Zupp, and in 1926, Wilson studied vaginal smears in swine. They found periodic variations in the cellular and leucocytic make-up of the smears, indicative of a cyclic rhythm.

A further step was made by Corner’s (’23) application of this method to the Primates. His observations on Macacus rhesus, though not entirely in line with the smear ﬁndings in the rodents, yet revealed a rhythm in the vaginal reactions. The average length of cycle in the Macacus, when the regular cycles are considered, is about 27 days, which is almost the same as in the human. During menstruation, which lasted 4 to 6 days, the vaginal smear contained erythrocytes, epithelial cells, and leucocytes. In the ﬁrst half of the intermenstrual interval there were relatively few epithelial cells and many leucocytes. About the middle of the interval a sharp drop in leucocytes occurred or even total disappearance. Leucocytes sometimes reappeared a few days later or were absent until onset of next menstruation or a few days before. During the second half of the intermenstrual interval there was an increased desquamation of epithelial cells. The pre-menstrual smear seemed to be thick and caseous, whereas the postmenstrual smear was rather thin and scanty.

These observations, though not establishing a deﬁnite succession of clear-cut stages, reveal the existence of a rhythm expressed mainly in the periodical increase and decrease in the number of the leucocytes and the epithelial cells. As Corner concludes:

There is to some degree a cycle of the vaginal secretion in this species. There is hardly enough evidence to warrant a correlation with the much sharper cycle of the rodents; moreover there was seen in the monkeys no massive desquamation of completely corniﬁed epithelial cells and no swarming of leucocytes into the epithelial debris. However, . . . . it seems very likely that the disappearance, or diminution in number of vaginal leucocytes, which usually happened about the tenth to the ﬁfteenth day before the onset of menstruation (in regular cycles of 25 to 30 days) is to be compared with the disappearance of leucocytes from the vagina of rodents at oestrus or shortly before the moment of ovulation.

In a later paper, 1927, in which the presence of menstruation without ovulation in Macacus rhesus was reported, Corner stated that daily vaginal smears, taken from animals with ovulative and non-ovulative cycles, were practically alike. The presence or absence of ovulation could not be ascertained by the examination of vaginal smears.

Allen, in 1927, also studied vaginal smears of the monkey, Macacus rhesus, and found varying numbers of epithelial cells in different stages of corniﬁcation. Some of these cells were quite normal, others were ﬂattened and their nuclei were pyknotic. The epithelial elements were present in greatest numbers during the latter half of the second and the whole third week of the cycle. Completely corniﬁed, non-nucleated cells frequently also appeared at these times. Leucocytes were present in greatest numbers before, during, and after menstruation and in least numbers or absent between the tenth and the twentieth to twenty-fourth day of the shorter cycle. During menstruation varying numbers of erythrocytes were present.

Allen’s ﬁndings are more or less in line with Corner’s observations, especially in regard to leucocytes. Corner found many leucocytes during and after menstruation and Allen before, during, and after. At about the tenth day both noticed a progressive diminution in the number of the leucocytes. This diminution lasted for a few days or almost up to the onset of next menstruation. In regard to ovulation, they agree that, whenever present, it occurs in the mid-period, between the tenth and the ﬁfteenth days.

The preparation of vaginal smears from the human has been in usein pathology for a long time for the study of various conditions, especially of bacterial infections. However, the application of smears to morphological and physiological studies has been extremely limited up to the last decade.

As early as 1847, Pouchet, in his book on “Ovulation and other related phenomena,” gave a description of human vaginal smears, which is chieﬂy interesting from a historic standpoint. Though unaided by modern technical methods, he was able to recognize the existence of a rhythmical reaction in the vaginal secretion. His work, however, was largely lost sight of and in no way stimulated attention upon the value of the vaginal content as an indicator in analyzing the phases of the sexual cycle. The recent interest and activity in these studies can in no sense be connected with or attributed to this early pioneer effort by Pouchet.

He describes the vaginal mucus as becoming less dense shortly before menstruation and as acquiring a peculiar odor (une odeur ‘sui generis’), to which he attributed an exciting effect upon sex desire. Microscopically, the vaginal ﬂuid showed fragments of epithelium and lacerated pieces, some consisting only of the ‘tubercule central’ (he evidently meant the ‘nucleus’), also large numbers of ‘globules muqueux’ (meaning probably the leucocytes) and some erythrocytes.

During the menstrual phase (‘période d’état’) he records the enormous quantity of erythrocytes, of mucous globules, and of small and transparent epithelial fragments. He believed that ovulation occurs toward the end of the menstrual phase, when the sex desire is most imperative. The menstrual phase is followed by a ‘period of desquamation,’ lasting approximately 10 days. This period is characterized by the detachment of a considerable quantity. of epithelial plates (‘Plaques d ’épithélium ’) .

During the sixth and seventh days after the end of menstruation (eleventh to twelfth day after onset) the vaginal mucus begins to lose its transparency and becomes heavier. Large numbers of epithelial plates are present and the mucous globules become more abundant. At about this time, or on the eighth day after menstruation (thirteenth day since the onset), some women experienced a feeling of heaviness or sharp pains lasting 1 to 2 days. Pouchet attributed these to contractions of the fallopian tubes, and not to the ovulative process, which he thought to occur much earlier.

Between the tenth and ﬁftenth days after cessation (ﬁfteenth to twentieth day after onset of menstruation) pieces of uterine decidua (‘ﬂocon membraneux’) were expelled in a number of cases. Pouchet interpreted this as an abortive process and believed that a woman could only conceive between the time of menstruation and the spontaneous fall of this decidua.

In 1921, Lehmann made a study of the diagnostic value of the human vaginal smear. Interested chieﬂy from a pathological and diagnostic point of view, he did not attempt to establish deﬁnite morphological and physiological relations between the changes in the vaginal ﬂuid and the ovarian and uterine cycles. He recognized, however, the dependence of certain vaginal conditions, such as secretion of glycogen, acidity, or bacterial growth on ovarian and uterine functions.

In 1925, I announced in a preliminary report some of the early results of my human vaginal smear studies. I held that there are deﬁnite morphological changes in the vaginal ﬂuid by which a diagnosis of certain physiological and pathological conditions is made possible.

Pregnancy, cystic or other degenerative changes of the ovaries, inﬂammatory processes, growth, etc., affect the entire genital tract, including the vagina, in a way which produces deﬁnite and typical changes in the consistency and make-up of the vaginal smear. The presence or absence of different types of desquamated cells, as well as the varying form and number of leucocytes, lymphocytes, erythrocytes and bacteria, offer a variety of criteria upon which a diagnosis of certain. conditions may be based.

These observations were mainly on the morphological changes of the various constituents of the smear, and not merely on quantitative estimates of the relative number of leucocytes or cells.

In the same year, Allen (’25) reported tests on human vaginal smears from gynecological patients in collaboration with Dr. Q. Newell. He stated that—

Although they found a decided variation in the number of leucocytes and epithelial cells at different times in the cycle, results were not nearly so clear-cut as in rats, because normally no corniﬁcation occurs in the vaginal epithelium of women. Furthermore, the smear test is of greatest value in animals, in which sexual changes are not as clearly marked externally. In the primates menstruation furnishes such a prominent milestone that vaginal smears seem of secondary importance for diagnosis.

A year later, King ( ’26) published her studies on human vaginal smears. She found that “in general the secretion is more scanty during the ﬁrst few days following menstruation and the cells show less degeneration during the first part of the cycle than in the late intermenstrual and premenstrual phases. The leucocytes are of the polymorph type, although an occasional mononuclear can be seen.” She also found a deﬁnite decrease in the number of leucocytes and a relatively high content of epithelial cells in one cycle of case D, on the thirteenth day after one onset and 16 days before the next. A similar condition was seen in the third interval of case F, beginning in the middle of a 22-day cycle.” She considers this fall in leucocytes as being probably comparable to the intermenstrual decrease noticed in some of Corner’s monkeys which may bear a relation to the time of ovulation. She states, however, that there were great irregularities in this respect in other cases.

Two types of secretion described by Lehmann for normal adults may be found, according to her, in the same individual and, in addition, there is a type in which the content in both leucocytes and epithelial cells is high. She concludes that—

The cellular content of the vaginal secretion of the normal human female is exceedingly variable and is a doubtful index of changes transpiring in the ovary and uterus. The indications of periodic variation are even less evident than those found by Corner for the monkey. Considering the higher degree of specialization, this is the condition which might reasonably be expected.

A Mexican gynecologist, Ramirez^[1] published, in 1928, a study of the human vaginal smear cycle. He recognized five provisional types of cells: I) Cells with a large and round nucleus clearly outlined, ﬁnely reticulated, and taking a violet-reddish color with Leischmann’s stain; cytoplasm reticulated, pale-blue, with or without inclusions. II) Cells with a large round or oval nucleus having a thick reticulum, granular or semigranular; cytoplasm spongy, bluish, with inclusions and vacuoles. III) Cells with a much smaller nucleus having a thick reticulum or forming granules, irregularly outlined, taking a blue-violet color; cytoplasm condensed, bluish, vacuolated with inclusions. IV) Cells with a small, pyknotic granular or simply structureless nucleus of bluish or wine shade; cytoplasm spongio-granular with abundant inclusions and with or without large vacuoles. V) Anucleate cells and cells with only a trace of a nucleus. The nucleus appears as a mass of granules which seem to be dispersed in the granular cytoplasm. All the granulations, nuclear or cytoplasmic, display the same bluish color.

Ramirez found that during menstruation the prevailing cellular type is: I (49 per cent), II (27 per cent), III (20 percent), IV (4 per cent), and V (0); leucocytes, scarce at first, increase toward the end. During the post menstrual period (2 to 3 days after menstruation) the cell percentages were as follows: I (1 per cent), II (9 per cent), III (24 per cent), IV (60 per cent), V (6 per cent) ; many polynuclear neutrophilic leucocytes. During the early interval (4 to 8 days after menstruation): I (0), II (15 per cent), III (25 per cent), IV (52 per cent), and V (8 per cent). During the interval (8 days after menstruation) : I (0), II (1 per cent), III (42 per cent), IV (45 per cent), and V (12 per cent) ; leucocytes exceptional.

Type V was prevalent in pregnancy, where the cellular percentages were as follows: I (0), II (0), III (18 per cent), IV (12 per cent), and V (70 per cent). The aspect of the smear was entirely different during pregnancy and the cells showed a tendency to form large groups or laminae. Ramirez arrived at the same conclusion that I had in 1925, that pregnancy may be diagnosed through the cytological characteristics of vaginal smears. His cellular classiﬁcation is, however, different from mine and his diagnosis is based rather on quantitative estimates. The increase of leucocytes during and immediately after menstruation and their decrease during the interval, as observed by Ramirez, are in accordance with the previous observations of Corner and Allen in monkeys and of King in the human being.

↑ A preliminary report on the same subject was published by this author in 1922, but unfortunately I have been unable to obtain the article.

Hartman, in his paper on gestation in Macacus rhesus (’28), presented some data on vaginal smear changes under normal conditions and during pregnancy. He found that the only successful mating in his Macacus took place at the ninth to twelfth day. During the fertile mating leucocytes were still present in the Vagina (though on the decline). Soon after copulation there was a temporary vaginal leucocytosis. He recognized two different types of leucocytes: those with unstained nuclei and those whose nuclei stained intensively with methylene blue. He stated that—

The total number of leucocytes increases during the second half of intermenstruum, reaching a maximum just before (or during) menstruation, whereas they go down to nearly zero in the mid-interval; the curve of greatest desquamation from the vaginal wall rises to a maximum in the latter part of the interval, to fall, usually very low, about the time of menstruation.

This is in accord with Corner’s and Allen’s views. He also made observations on the cellular changes during gestation. He noticed the presence of the cells of Papanicolaou (navicular). He also found erythrocytes in the pregnancy smear for about 23 days, beginning at the twenty-sixth day since the onset of last menstruation (or 14; to 17 days after fertile copulation). He compares this with the bleeding observed by Long and Evans at the fourteenth day of gestation in the rat. In later stages of pregnancy Hartman found a leucocytic decrease and an increase of the resorption of the placental hormone as revealed by the exacerbation of the sex color. From these observations, he believes that “the disappearance of leucocytes from the vaginal smear is due to the follicular hormone.”

In the same year a detailed cytological study of the human vaginal secretion was published by Moser (’28). She studied ten cases of virgins and one sexually abstaining woman. She examined vaginal smears daily or every second day during one or two menstrual periods. The ﬂuid was taken with the aid of a ‘platinum loop’ and the smears were dried and then ﬁxed with methyl alcohol and stained with Delaﬁeld’s haematoxylin and eosin. The results were classiﬁed according to the relative number of epithelial cells, leucocytes, corniﬁed cells (Schollen), and detritus.

Moser was unable to detect any regularity in the consistency of the vaginal ﬂuid during the successive phases of the menstrual cycle. Epithelial cells were always present. Their cytoplasm and nucleus showed various degrees of degeneration. The whole structure was often covered by abundant bacteria. The nuclei of the cells were degenerated, pyknotic, or fragmented.

Isolated eosinophilic cells (Schollen) were found. Only in one case were they very numerous (about 7 per cent of the total). Moser also described nucleated cells which were stained darker with eosin and which could be interpreted as transitional forms. She found no evidence that the human vaginal epithelium undergoes a periodical corniﬁcation. There was great fluctuation in the number of leucocytes, but no typical leucopenia. In one case there was a premenstrual increase in the number of leucocytes, but this was not constant. Most of the leucocytes were polymorphonuclear, some enclosed within the cells. Mononuclears were also occasionally observed. Most of the leucocytes were in process of degeneration. Her conclusion was that the morphological changes of the vaginal secretion are not typical enough to supply information regarding the functional condition of the ovaries.

A comparative review of all the investigations of this character conducted in various classes of mammals shows that the results obtained in the rodents are by far the most satisfactory. The vaginal smear changes in these animals are clear-cut and characteristic of the more fundamental changes occurring in the ovary and the uterus. The time of ovulation can be accurately diagnosed in the rodents by deﬁnite smear syndromes. The moments in their sexual rhythm may be easily" established by the microscopical examination of a sample of vaginal ﬂuid. This has greatly advanced our knowledge of the sexual and reproductive phenomena in mammals, and endowed us with indicators which may be profitably applied for obtaining valuable experimental information. Various problems related to the ovarian hormones in particular as well as other hormonal reactions can now be attacked with greater accuracy.

In the higher mammals and in the Primates the results obtained so far have not been of equal clearness. In monkeys and in the human female, in which the sexual rhythm is revealed by the menstrual bleeding, Very little information has been gained by the application of the vaginal smear method. We have learned that Primates, as Well as other mammals, possess a deﬁnite vaginal rhythm. It has also been observed that certain physiological and pathological conditions of the female genital tract are revealed by vagi.n.al discrepancies. These facts fully justify continued researches in this ﬁeld, and it is believed that the present contribution furnishes much additional material toward a clearer understanding of the human cyclic changes as evidenced by an analysis of the vaginal smears of women.

2. Application of the Vaginal Smear Method in the Human

Material and technique

The correlation of the vaginal smear stages, as described in the rodents and in various other mammals, with analogous stages in women offers great difﬁculties. The extreme prolongation of the human female sex cycle has had a modifying effect upon the manifestation and the sequence of the vaginal ﬂuid changes. The various stages are less typically expressed than in some of the lower mammals and are subject to a greater variability.

During the ﬁrst 4 years of my studies (1920 to 1924), I was unable to correlate my observations on the human with those obtained in other mammals and more particularly in the rodents. In the spring of 1924, however, I ﬁrst appreciated a stage interpreted as a transition between the copulative and the ovulative stage. And yet both these stages, and especially the second one, looked quite different from the corresponding stages in the rodents.

An explanation for these differences was offered one year later, when I ﬁrst examined a human post-partum smear. This type of smear showed a surprising similarity to the missing stage III, which, in the guinea pig, is Very closely and typically related to ovulation. It thus became apparent that the guinea pig stage III was not primarily an ovulative one. Its almost infallible relation to ovulation was a mere coincidence. It should be interpreted as resulting from utero-vaginal catabolic processes, corresponding to a pseudo-postpartum condition. It also became evident, from a morphological standpoint, that the prooestrous hyperplasia should be interpreted as a pseudopregnant condition, and the onset of the cyclic desquamatory and catabolic phenomena of the utero-vaginal tract as the equivalent of a pseudoparturient or pseudoabortive process.

If this explanation were correct, it would necessarily lead to the old assumption that human menstruation and its associated catabolic phenomena are of a pseudo parturient or pseudoabortive character (Beard, ’97 ). lonsequently, a pseudo-postpartum stage, corresponding to the guinea pig stage III, ought to appear in the human toward the end of the menstrual process. This has been fully corroborated by later ﬁndings. Toward the end of menstruation a pseudo—III or pseudo—postpartum stage has been identiﬁed marking the end of the destructive period, just as the guinea pig stage III marked the end of the utero-vaginal destructive phenomena. A positive explanation has thus been offered for the significance of certain smear stages, and their classiﬁcation may be thus based on a fundamental interpretation of the female sex-cycle processes.

Most of the smears which I examined up to February, 1925, were obtained from a single human case (special case). This was an ideal case to work with, because of perfect regularity in the menstrual periodicity and of a complete lack of any serious disturbance or of any bacterial contamination. It has been by far the most normal and typical case that I have had an opporunity to study. Several smears obtained during this ﬁrst period from the clinic of Cornell Medical College in New York City, through the courtesy of Dr. John McGrath, were taken mostly from pathological cases. These have been little utilized for the study of normal standards.

In February, 1925, a valuable association was established with the Woman’s Hospital of New York City. This was rendered possible through the cooperation of Dr. George Gray Ward, director of the Woman’s Hospital, Dr. Charles R. Stockard, administering a grant from the Committee on Problems of Sex of the National Research Fund, and Dr. Robert L. Dickinson, representing the Maternal Health Committee. A special assistant was assigned to this work, whose technical skill in the preparation of smears has been extremely valuable.

Our ﬁrst effort was to obtain a. sufficient number of control smears from normal cases. We selected twelve healthy young women, the majority from the personnel of the hospital, and arranged to obtain daily vaginal smears through several menstrual periods without interruption. Later, the material was extended to include smears from pregnant women who came to the clinic of the hospital for examination, and also from various pathological cases treated at the hospital.

From the start of this Work at the Woman’s Hospital up to the end of 1929, more than 900 cases were examined. Almost every important pathological condition has been more or less thoroughly considered. There have been smears from cases of pregnancy, abortions, threatened abortions, ectopic pregnancies, various infections or inﬂammatory processes of the genital tract, cystic ovaries, pus tubes, and various benign and malignant tumors of the ovaries, uterus, cervix, vagina, and mammary glands. There were also several cases of sterility due to various causes.

This study of pathological smears has helped in an understanding of normal processes and opened a new possibility for diagnosing various gynecological conditions. In a preliminary note (1925), I referred to the application of the vaginal smear method as a guide in the diagnosis of pregnancy and of pathological conditions.

In the last few years attention has been given to the characteristic smear changes Which seem to accompany cases of malignant tumors of the genital tract. A preliminary report on these observations was given at the “Third Race Betterment Conference” (1928).^[1] Several types of abnormal cells with enlarged, deformed, or hyperchromatic nuclei are present in such smears.‘ Leucocytes are numerous and display a high phagocytic activity. Mononuclears become especially prominent. Erythrocytes are commonly present. A more detailed discussion of this phase of the study will be given in a subsequent publication. In this paper We intend to describe only the normal smear ﬁndings and to discuss their possible application in a detailed analysis of the human female sex cycle.

↑ Unfortunately, many stenographic errors have been included in this article, some altering the meaning of the text.

The technique used in these studies has been rather simple. In dealing with human material one beneﬁts greatly by simplifying methods and avoiding unnecessary technical difficulties. In order to obtain the samples of vaginal fluid a glass pipette 8 inches long and 3/8 of an inch in diameter was used. The pipette was slightly curved at the end. The opening was small and a strong rubber bulb at the opposite end was used to produce suction in collecting the ﬂuid (ﬁg. 1). The sample of vaginal ﬂuid may be taken with the aid of one of these pipettes by a technician or a nurse. Some of these patients were able to take their own smears. The sample is obtained from a region5 or ‘6 cm. within the vagina and is spread on a glass slide as a heavy ﬁlm, much heavier than an ordinary blood smear. Soon after the spreading and before the smear has dried the slide is placed in a jar containing equal parts of 95 per cent alcohol and ether for quick fixation. When ether was not available, plain alcohol was used.

Fig. 1 Pipette used for taking samples of human vaginal fluid.

Proper ﬁxation of the vaginal smears before drying is important, although this has been somewhat underestimated by most of the workers in this ﬁeld. A smear which has been allowed to dry either before or after ﬁxation may serve for the simple recognition of a typical stage, as has been done on rodents, but is not suitable for a detailed cytological study, especially in the human. The cells, as well as all other elements, are affected by the drying and lose their true appearance. Several investigators have probably failed to recognize certain morphological characteristics because of the use of dry smears.

The slides were kept in the alcohol-ether ﬂuid for from half an hour to one hour, but smears may be kept in this ﬂuid for several hours or even days without much harm. After fixing in this way the slides are carried through 80 per cent, 70 percent, and 50 per cent alcohol into distilled water as preliminary to staining.

The staining of vaginal smears with haematoxylin-eosin gives good general results and a fair differentiation of the corniﬁed cells, and such a stain may be considered as very suitable for the recognition and study of the fundamental morphological changes in the smear of rodents. This stain, however, in the study of human smears, did not give satisfactory results.

After using various staining ﬂuids, I succeeded in developing a simple stain, consisting of a combination of haematoxylin, eosin, and waterblue, which seems to offer certain advantages for the general staining of vaginal smears. This secures a sharp outline of the various cell types with a variety of shades from an intense blue to an eosin—red. The corniﬁed cells, which are almost constantly present in the vagina, take an intense eosin color and are sharply differentiated. On the other hand, cells derived from the deeper vaginal layers, which are free from corniﬁcation, are stained strongly blue. Partly corniﬁed cells show various shades of purple. Superficial secretory cells, containing mucus, display a typically characteristic bluish—purple tone. Leucocytes, as well as bacteria, are also sharply outlined. Eosinophiles show distinctly red granulation. Mucus takes a bluish or a purplish shade. The contrasts are much sharper than with ordinary haematoxylin-eosin and the corniﬁed cells are particularly differentiated. This is important for the Vagina, in which the cornification is Very pronounced, especially during certain periods of the cycle.

The exact method is as follows: Slides are ﬁrst stained with Ehrlich’s haematoxylin for about 5 minutes, then carried into running water for at least a quarter of an hour. After this they are stained with a 0.5 per cent eosin for about 3 to 4 minutes, rinsed well in water, and ﬁnally stained for 1 minute in a 0.5 per cent solution of waterblue. Following this they are rinsed well in water and then carried through 50 per cent, 70 per cent, 80 per cent, 95 per cent, and absolute alcohol into xylol and mounted in Canada balsam. The stain is permanent and does not fade easily.

Several other methods have been used, but none was as satisfactory as the above. Many of the slides have been stained with speciﬁc bacterial stains, especially with Gram’s, for the study of the bacterial ﬂora.

The twelve selected cases from which daily smears were obtained were all normal and in good health. Racially, eight were Americans, two Irish, one British, and one Italian. Ages ranged from 19 to 28 years, one, however, being 35 years old. Five were unmarried and seven, though previously married, were not living with their husbands. They were of average height and their weights ranged from 100 to 152 pounds. Smears were taken daily, except on holidays, for 2 to 3 months, beginning February 9, 1925.

In addition to these, many smears have been examined from other cases. The largest number has been taken from one individual (our special case), covering about ﬁfteen periods and studied at different times since 1920. A number of smears was also obtained from sixteen operated cases at the Woman’s Hospital. These were patients with slight pathological disturbances, in whom laparotomy was necessary. Smears were taken before and after operation, and the ovaries were examined during the operation in order to ascertain the relation of the various smear stages to the condition of the ovaries and especially to ovulation. This has been rendered possible by the helpful cooperation of the director of the Woman’s Hospital, Dr. George Gray Ward.

The total number of smears studied is 1022, arranged as follows:

First day of menstruation: 39 smears (13 from the special case and 26 from the other cases); second day: 39 (13 + 26); third day: -39 (13+26); fourth day: 39 (12+27);

Plates

Plate 8

Photomicrographs of human vaginal smears colored by hand as in stained smears, in order to show the relative number of cornified cells (stained red).

32 Twelfth day. Typical copulative smear with leucopenia. Same as ﬁgure 13.

33 Ovulative type of smear. Fourteenth day. Oc. 3,0bj. 8, B. 19.

34 Fifteenth day. Post-ovulative smear with continued high corniﬁcation and Ieueopenia. Same as ﬁgure 17.

35 Twenty-third day. Typical premenstrual smear. Same specimen as in ﬁgure 18. 0c. 6, Obj. 16, B. 23.

Plate 9

Drawings of various types of cells found in normal human vaginal smears.

36 to 60 Cells from human vaginal smears at different stages of the normal menstrual cycle.

Plate 10