Talk:Paper - Human ova from large follicles - including a search for maturation divisions and observations on atresia

HUMAN OVA FROM LARGE FOLLICLES; INCLUDING A SEARCH FOR MATURATION DIVISIONS AND OBSERVATIONS ON ATRESIA

EDGAR ALLEN, J. P. PRATT, Q. U. NEWELL, AND L. J. BLAND

University of Missouri, Henry Ford Hospital, and Washington University School of Medicine*

FIVE PLATES (THIRTY FIGURES)

INTRODUCTION

In examining human ovaries at operation one often wonders how many of the larger follicles present may be normal and which one may be destined to rupture at the next ovulation. These follicles may range in size from 3 or 4 to 15 mm. in diameter or even larger and still retain a normal appearance upon gross examination. The number may vary between two or three and thirty-five in a single ovary.

Another question which has long been a matter of discussion is the maximum size which the normal follicle in the human ovary may attain before rupture at ovulation. Thomson (’20), from his studies of human ovaries, believes that follicles larger than 5 mm. in diameter are probably not normal. Gynecologists are often inclined to consider follicles larger than 15 mm. as probably cystic. Judging from observations in other mammals, the pre-ovulation size of the human follicle probably is not greater than that of the fully formed corpus luteum. In woman this seldom exceeds 20 mm. in greatest diameter.

This work was a cooperative effort utilizing material from the gynecological services at Henry Ford Hospital, Detroit, and Barnes Hospital, St. Louis, The work done in St. Louis was carried on during the summer of 1928 in the Department of Surgery and Gynecology and during the summer of 1929 in the Department of Obstetrics and Gynecology of Washington University School of Medicine. It was further supported by grants from Drs. Evarts A. Graham and Otto Schwarz.

1

THE AMERICAN JOURNAL OF ANATOMY, VOL, 46, NO. 1 JULY, 1930 2 EDGAR ALLEN ET AL.

With these points in mind, it was decided to make a careful census of the number, size, and condition of follicles in ovaries at varying times in the menstrual cycle. This study was carried on in conjunction with the search for human tubal ova? and analyses of hormone content of ovarian tissues. It also provided a diagnosis of normality or atresia of follicles which was of interest when correlated with their hormone content.

Another major objective of this study was a search for maturation stages in human oocytes. To date there have been few descriptions of such stages and in those which have already been reported several were clearly concerned with atretic follicles. Also in a majority of cases the stage of the menstrual cycle at which the ovary was removed was not known. Some knowledge of the menstrual history is necessary in judging normality of large follicles if the final eriterion is considered the possibility of their undergoing ovulation. From what is known of the relation of maturation to ovulation in the majority of other mammals, maturation divisions in normal ova should be sought for just before ovulation time—probably less than ten or twelve hours before rupture of the follicle is due.

During the past fifteen years studies of recent corpora lutea have shown that ovulation in woman usually occurs at some time in the intermenstrum (Schréder, °15, and others). This has recently been further confirmed and the time of ovulation in the menstrual cycle more definitely located by the recovery of unfertilized ova from the uterine tubes of woman (Allen, Pratt, Newell, and Bland, ’28, ’29) and of the monkey (Corner, ’23, and Allen, ’27 and ’28). In a majority of these cases ovulation occurred on or shortly before the fourteenth (morning of the fifteenth) day following the onset of the previous menses. Therefore, in searching for maturation stages, although ovaries removed at all stages of the menstrual cycle were studied, especially careful examination

7In press: Carnegie Publications, Contributions to Embryology. 3In press: Amer. Jour. Physiol. HUMAN OVA FROM LARGE FOLLICLES 3

was made of those removed between the twelfth and sixteenth day of the cycle. So far, this search has resulted in the recovery of one ovum in the process of maturation division.

Many of the larger follicles visible on the surface of the ovary, although apparently normal on gross examination, proved on study of the contained ova to be atretic. Much of the work previously done on follicular atresia in the human ovary has been restricted to the smaller follicles. Consequently, several interesting observations on varying degrees of early atresia in large follicles have been made in this material.

LITERATURE

During the hundred years since Von Baer (1827) recovered the first ovum from the graafian follicle,t a great deal of work has been done on ovaries of mammals. It is not possible to adequately review this earlier work in this paper, but a few of the outstanding contributions should be mentioned. The works of Barry (’38 and ’39) on the rabbit and Bischoff (’42 to ’52) on the dog and guinea-pig were especially significant. Later came the contributions of Waldeyer (’70), Van Beneden (’75), Hertwig (’75, ’78), Flemming (’85), and Nagel (’88) which made use of many improvements in technique. Schottlaender (’93), Henneguy (’94), Stoeckel (’98), Honore (’99-’00), and Von Ebner (’00) have contributed important papers.

Some of the best work on mammalian ovaries has been done on the smaller ‘laboratory animals.’ Late growth, maturation, and fertilization have been described, in the mouse and rat by Sobotta (’95), Kirkham (’07), Kirkham and Burr (713), Lams and Doorme (’08), Sobotta and Burckhard (’10), and Long and Mark (711). The works of Rubaschkin (’05), Heape (’05), Van der Stricht (’08, ’23), and Longley (’11) have given us descriptions of maturation in the guinea-pig, rabbit, bat, dog, and cat. Further studies of reproduction in the guinea-pig with emphasis on cyclical changes have been

For a short review of Von Baer’s work, see Addison, ’27. 4 EDGAR ALLEN ET AL.

contributed by Loeb (’11, °14), Walsh (’17), Stockard and Papanicolaou (717), in the rat by Long and Evans (’19-’22), in the opossum by Hartman (716, ’23), in the pig by Corner (717) and McKenzie (’26), in the ferret by Robinson (719), in the mouse by Allen (’23) and Parkes (’26), and in the rabbit and cow by Hammond (’23) and others.

The study of the ovarian ovum of man up to the time of Nagel’s work had been limited to autopsy material. Nagel (’88) studied ovaries removed at operation by opening the follicles and recovering the ova from the follicular fluid. His ovarium ovum described in 1888 is still used in several texts (Keibel and Mall, 710, and others) to illustrate the nearly mature human ovum, but since the nucleus was intact and centrally located, it is clear that this ovum had not begun maturation divisions. A careful search of the literature on human. ovaries shows that most of the studies were concerned with relatively small follicles. The work on large follicles seems to be restricted for the most part to the histology of the walls. As far as the writers are aware, only three observations on maturation stages in human ovarian ova have been reported.

Maturation stages in the human ovum

Thomson (’19) deals with maturation of the human ovum. His is pioneer work in the difficult search for maturation stages in ovaries of women. His work is open to several severe criticisms. The specimens were all obtained from autopsy material and were not timed with regard to the menstrual cycle. None of the follicles described were larger than 5 mm. in diameter, and the author considers that follicles in the human ovary larger than this are abnormal. ‘Polar bodies’ are described in ova which also contain clearly defined, centrally located nuclei. These ‘polar bodies’ in several cases are no larger than follicle cells. In several specimens follicle cells of the cumuli are pycnotic—a certain indication of atresia. The dense cumuli surrounding one or two of the ova described as undergoing maturation indicate HUMAN OVA FROM LARGE FOLLICLES 5

clearly that these specimens are immature ova. The dispersed cumuli of others are clearly indicative of atresia. In several cases corona-radiata formation is described where figures show none. In several ova doubtful structures are interpreted as first and second polar bodies and the conelusion is drawn that in man both the maturation divisions are completed in the ovary before ovulation. Thomson’s descriptions are clouded by many questionable assumptions. It seems doubtful if any of the ova described by him can be considered as representing normal maturation stages.

Thomson (’20), in discussing the ‘ripe human Graafian follicle,’ makes the statement, ‘‘ovulation appears to take place at any stage in the menstrual cycle’’—-an opinion formed from altogether insufficient evidence. An opinion that the granulosa cells are stripped from the wall of the follicle at ovulation is apparently derived from shrinkage artefacts in atretic follicles. Several other points brought out in this paper are open to severe criticism.

Dixon (’27) described a single section of an ovarian ovum in which the first polar body and the metaphase of the second maturation division had been formed. This ovum was in a flattened follicle compressed between two larger follicles. Its measurements after fixation were 0.10 to 0.13 mm. in outside diameter ; vitellus, 0.08 to 0.09 mm. The ovum is pictured as surrounded on all but one side (where the follicle cells are reduced to one layer) by a rather dense cumulus. There is no evidence of corona-radiata formation. Dixon accepts Thomson’s interpretation that both polar bodies are probably formed in the human ovum before ovulation.

Hoadley and Simons (’28) described a beautiful spindle in the late metaphase or very early anaphase in a human oocyte. The spindle fibers are clearly defined and most of the chromosomes (sixteen or eighteen) are in the equatorial plane, although one or two had started migration toward the poles. From the tetrad formation of some of the chromosomes this was considered to be a first meiotic division. It is probably the best specimen so far obtained of maturation 6 EDGAR ALLEN ET AL.

of the human ovum. As the authors note, however, the cumulus is partly dispersed on one side, and the nuclei of many of the follicle cells are pyenotic. Also the ovum is still attached to the follicle wall and the size of the follicle (as far as can be judged in the section) is less than 1 mm. in greatest diameter. No mitoses are to be found among the follicle cells and the follicle is indented by an adjacent smaller follicle. It seems clear from these considerations that this is an atretic follicle. A second oocyte in a follicle in more advanced atresia is also described.

Hinselmann (’27) reported failure to find maturation std@ges in human ova from more than one hundred large follicles. His explanation that maturation probably follows ovulation in man, as in the dog, contrasts with the opinion expressed by Thomson and by Dixon. Our definite identification of first polar bodies and chromosomes of second maturation spindles in several unfertilized tubal ova brings further evidence to bear upon this point.

From a consideration of these papers it is evident that if maturation stages are to be found in normal human ova the attack must be focused upon one or two of the largest follicles in ovaries obtained just before the time of expected ovulation.

Atresia in ovarian follicles

The literature on atretic follicles is very extensive, and consequently only the points bearing upon the findings in the present study will be mentioned here. A recent paper by Branca (’25) which deals with atresia in the rabbit, guineapig, and mouse gives a lengthy bibliography.

For the most part, diagnosis of atresia in large follicles of the human ovary has depended largely on the condition of the granulosa and theca layers of the follicle walls—chiefly because routine test sections rather than serial sections have constituted the greater part of the material studied.

Waldeyer (’70) and Henneguy (’94) have shown that many ova are eliminated from the ovary, only a very small proportion of the total number attaining maturity. Schottlaender HUMAN OVA FROM LARGE FOLLICLES 7

(’91) studied follicular atresia especially in rodents and in 1893 extended his work to human ovaries. He dealt chiefly with small follicles in fetal ovaries.

Henneguy (’94) described the entry of follicle cells and possibly leucocytes into ova during the late stages of degeneration and interpreted their action as phagocytic.

Stevens (’04) described atresia of small follicles in human ovaries during premenstrual life. He noted follicle cells penetrating ova which were in a degenerate condition. Seitz (’05-’06) examined ovaries of thirty-six pregnant women without finding follicles approaching a ripe condition.

Loeb (’11) and Asami (’20) studied follicular atresia in the guinea-pig and rabbit, respectively. The former concluded that following ovulation in the guinea-pig many of the large and medium-sized follicles became atretic. Asami was unable to detect any cyclic variation of follicular atresia in the rabbit. Clark (’23) also described atresia in the guinea-pig ovary. The difficulty of recognizing early atretic changes was discussed and the conclusion reached that often degenerative changes in follicles first become apparent in the contained ova.

Athias (’09) described several types of atresia, including fragmentation and invasion of the ovum by follicle cells. Evans (’16) used vital dyes to distinguish early atresia of follicles. He described the entrance of macrophages into the egg.

Shaw (’25) discussed ‘‘the fate of the Graafian follicle in the human ovary.’’ He recorded eight or ten ‘ripe follicles’ in the postmenstrual phase of the cycle, but did not mention the sizes of these follicles, and ‘ripeness’ and normality or atresia are apparently judged only by sections of the follicle walls. He cited three cases in a series of fifty where two corpora lutea in similar stages of development indicated ovulation of twin ova.

Arnold (’12) described two very unusual ovaries from an eighteen-year-old negress. These ovaries contained many polyovular follicles, there being eighty-eight follicles containing more than one ovum. Liquor folliculi had begun to 8 EDGAR ALLEN ET AL.

form in forty-five of these follicles, but the largest was quite small (only 0.633 mm. in diameter). The contained ova ranged from 50 to 180 in diameter and the author considered that some of the smaller ones were being eliminated.

Hartman (’26) thoroughly reviewed the question of polyovular follicles and also added some new evidence. He believes that all but one of the contained ova are usually eliminated before the polyovular follicle grows to any considerable size.

Allen (’28 a) described a medium-sized follicle in a monkey ovary which contained four ova in varying stages of atresia. Three were grouped at one pole of the follicle, and showed several interesting stages of progressive degeneration.

Possible parthenogenesis and fragmentation of ova

Loeb (’11) has described structures interpreted as parthenogenetic development of ovarian eggs in ovaries of the guinea-pig; Newman (’13) and Sansom (’20) described similar processes in atretic follicles in the armadillo and water vole, respectively. Kingery (’14) and Addison (717), after long searches in the mouse ovary, reported only fragmentation of ova.

Charlton (°17) found fragmentation and phagocytosis to be the fate of unfertilized mouse ova in the tubes. Smith (’25) describes similar degenerative changes in uterine eggs of the opossum. (In the opossum the transit of the ova through the tubes is very rapid.)

Quantitative studies of atresia

Henle (’73), Hyse (’97), Joessel (’99), and Hansemann (713) have attempted estimates of the number of ova in ovaries at different periods of life. Perhaps the most thorough study of this sort was that by Arai (’20), who chose the albino rat as a subject and dealt with a considerable number of ovaries throughout the whole life span. His results show the elimination of enormous numbers of ova. HUMAN OVA FROM LARGE FOLLICLES 9

Allen, Kountz, and Francis (’25) attempted an estimation of the elimination of follicles from the pig ovary during sexual maturity with especial reference to cyclic variation in this process. They found an elimination of many ova even in the later stages of growth of the follicle.

Engle (’27) has contributed a quantitative study of follicular atresia in the mouse and correlated his data with the oestrous cycle. He found the incidence of atresia greatest during the first day of the dioestrum. |

MATERIAL AND METHOD

This study attempts to correlate the following points: the time in the menstrual cycle, the stage of development of the most recent corpora lutea, the size, number, and condition of the larger follicles. In a study of this last point the condition of the ovum is, of course, the primary consideration, but the nature of its cumulus, whether attached to the follicle wall or floating free, the presence of a corona radiata, and the incidence of mitotic figures in both cumulus and granulosa also are important. Proliferation of the follicle cells, which has been shown in other mammals to continue up to the time of ovulation, served as a definite criterion of normal, growing follicles.

In the course of this work more than sixty ovaries or parts of ovaries have been studied. Many were removed from cases where tumors necessitated the removal of the uterus. At times one or two large follicles alone were excised without removing a whole ovary. In such cases follicles visible on the other ovary or the part of the ovary not removed were measured and tabulated. Specimens which were grossly pathological were not included.

As soon as possible after removal of the ovary at operation, the follicles were dissected out and measured. If the follicle was approximately spherical, a single average diameter was taken. If it was flattened or elongated, three diameters were recorded. Then a search was made through the wall of the intact follicle under a strong light for the 10 EDGAR ALLEN ET AL.

ovum. Sometimes the ovum could be seen through the follicle wall. More often it was not visible. Then a small incision was made on the surface of the follicle while under observation with the low power of the binocular. If the point of incision was supported and the follicle not allowed to collapse, the liquor folliculi would flow out slowly through the incision. Often the ovum could be seen to float up and through the incision and could be picked up immediately with a small pipette. More often it was necessary to cut further the follicle wall and remove the liquor folliculi. Usually a short search through the liquor was enough to locate the ovum. In cases where the ovum was still attached to the follicle wall by a persistent cumulus, the follicle wall was turned inside out and the ovum freed with a dissecting needle. Where possible, if the ovum was still attached to the follicle wall, the location of the point of attachment with regard to the surface of the ovary was noted. A small slip of the follicle wall was taken from a few specimens to study the condition of the granulosa layers.

In cases where the ovum was not surrounded by a dense cumulus, measurements were made while the ovum was still fresh in liquor folliculi. In other eases the surrounding follicle cells made accurate measurements difficult.

-The ova were fixed in Zenker’s or in Flemming’s fluid, or in Zenker-formol. The fixative precipitates large opaque masses from the liquor folliculi. Consequently, it is best to remove the ova to normal saline solution before fixation to prevent their loss in these opaque masses of coagulum. Ova were stained with eosin after fixation to aid observation during dehydration, clearing, and embedding. This staining makes it quite easy to keep the ovum in view up to the time of cooling the paraffin and it can be located again as a small red speck in the paraffin block. Ova were sectioned at a thickness of 10 and stained with iron-alum or with Delafield’s hematoxylin and counterstained with eosin or orange G.

This method of recovery of ova from large follicles and section of the ova and small pieces of the follicle wall separately HUMAN OVA FROM LARGE FOLLICLES 11

has several advantages over sectioning whole follicles. It also permits the use of the fresh liquor folliculi from the same follicles for hormone analyses by direct injection into spayed rats. As far as we are aware, there has been no previous attempt to correlate anatomical studies and endocrine tests of the same follicles.

OBSERVATIONS Distribution of ovaries with regard to the menstrual cycle

The ovaries studied were distributed through all stages of the menstrual cycle, but, since a search was being made for tubal ova and maturation stages in ovarian oocytes, a distinct effort to date the operation during the intermenstrum placed a large number of cases between the tenth and sixteenth days of the menstrual cycle. There were six cases on the fourteenth day, four each on the twelfth and thirteenth days, three each on the fifteenth and sixteenth days, and two each on the tenth and eleventh days of the cycle. Other times in the cycle were represented at daily intervals by one or sometimes two cases.

In addition, four ovaries were obtained during pregnancy ; three at full term and one at about three months’ gestation.

Distribution of ovaries as to age

The ages of patients from which one or both ovaries or a part of one ovary were obtained ranged from 20 to 44 years. The cases were distributed as follows: five at 26 years, three each at 30, 33, 34, 37, 38, and 40 years, and one or two each for ages of 20, 21, 24, 27, 28, 31, 32, 35, 41, and 44 years. Cases where ovaries showed signs of sclerosis or the patient’s history indicated approach of the menopause were discarded. It is realized that differences in ovaries dependent on age, especially as concerns the number of follicles present, may make comparisons difficult. 12 EDGAR ALLEN ET AL.

Number and size of the larger follicles

These data are listed in the fourth and fifth columns of table 1. The number of the larger follicles ranged from two or three to thirty-two and averaged about eight follicles per ovary. Although the follicles in mammalian ovaries are often not equally distributed between the two ovaries in individual cases, observations of large numbers of cases show that in dealing with averages the distribution is approximately equal in right and left ovaries. The average number of eight large follicles per ovary in this series might therefore be doubled to obtain an average of sixteen follicles per person.

The sizes of the follicles included in these data ranged from 3 mm. up to about 20 mm. in average diameter. Many of these follicles were shown after section of the ova to be definitely atretic.

Variation in size and number at different phases of the menstrual cycle

Large follicles (7 mm. or more in diameter) were present in ovaries of this series at nearly all stages of the menstrual cycle. Usually, however, ova from the larger follicles at times other than the early intermenstrum showed signs of atresia. The greatest number of follicles present seems to oceur during and just after a menstrual period. It is difficult, however, in this limited series, some specimens of which may have been involved with pathological conditions, to find a clear-cut variation which is peculiar to any special phase of the cycle.

The condition of the larger follicles

In looking at an ovary removed on the tenth or twelfth day of the menstrual cycle which might have from four to ten or twelve follicles, ranging in size from 2 or 3 to 10 mm. or more in diameter, visible on the surface one might tentatively expect that one of the largest follicles might undergo maturation and be ovulated within the course of the next two or HUMAN OVA FROM LARGE FOLLICLES 13

three days. The majority of the larger follicles probably would be eliminated by atretic processes. One might reasonably expect, in the absence of concrete data, that several of the smaller follicles 2 or 3 mm. in diameter would be the source of the ovum to be ovulated during the following cycle or, if ovulation alternated between the two ovaries, the second month thereafter. It has been of interest therefore to ascertain by cytological study the actual number of these follicles which were normal.

Probably the most reliable criterion of normality of the follicle is the presence of mitotic division of follicle cells of the cumulus or granulosa layers (figs. 4, 6, and 8). Where active proliferation of follicle cells is found, it is seldom possible to detect degenerative changes in the ovum. It has been shown in other mammals that division of follicle cells continues up to just before the time of ovulation; consequently, absence of mitoses from large follicles should stimulate a careful search for degenerative changes in the ova.

It is known in lower mammals that maturation divisions begin just before an ovum is due to be ovulated. Up to probably less than a day before ovulation, the nucleus should be centrally located in the ovum and the ovum attached to the wall of the follicle. Consequently, ova which are floating free in liquor folliculi found in large follicles at times other than the intermenstrum may reasonably be expected to have begun degeneration. :

From recent studies of tubal ova of monkeys and man w may expect the primate ovum at ovulation to be surrounded by from two to four layers of follicle cells. All of these points have been taken into consideration in judging normality or atresia of ova in large follicles as listed in column 6 of table 1.

As large apparently normal follicles are dissected from the ovary one notices that the theca layers often are edematous, the follicles appearing somewhat like white grapes after they have been squeezed from the skins. As observed under bright illumination the follicle is usually thin-walled and quite translucent. The liquor folliculi is clear and light-straw-colored. 14

EDGAR ALLEN ET AL.

TABLE 1

Number, size, and condition of the large follicles in human ovaries

DAY OF | wuMBER | USF sor | Mn Orang, routiclns | SPORES | nanxs Mm. 1P 26 2nd (| R.Ov., 32 1=10 11 of these ova/This ovary con5= 8 were section-| tained 1 cor2= 7 ed; all atretic| pus luteum. 9= 5 Left ovary 6= 4 also contained 9= 3 many follicles 2P 20 2nd L.Ov., 21 1= 8 6 ova sectioned; Ovary had 1 3= 7 all in ad-| corpus. One 6= 6 vanced atresia] 4-mm. follicle 38= 5 contained 4 8= 4 degenerating eggs 3P 5th 1 Ov., 7 = 9 ‘All 7 ova inlAll ova unat= 8 varying de-| tached to fol= 7 grees of atre-| licle walls 2= 6 sia l= 4 4P 7th 1 Ov., 5 =7 One 7-mm. fol= 6 licle contained i= 5 normal ovum; = 4 others atretic | 5P 8th 1 Ov., 5 1=10 4 of these ova|Not attached to 1= 7 in early atre-| follicle wall 2—= 6 sia l= 4 6P 32 9th R.Ov., 17| 1=16 12x11 |Largest follicleLeft ovary had 2= 7 had normal or; same number 4= 6 early atretic, of follicles 3= 5 ovum; 3 other 6= 4 normal follieles 1=7, 1== 5, 1=4 mm.; others atretic 7P | 30 | 5thpast, | R.Ov., 30 1=10 10-mm. follicle Ovary also conprobably 3= 7 | had normali tained 1 large 9th or 5= 6 / ovum; others} corpus luteum 10th 4= 5 | in varying de5= 4 grees of atre

12=3 or less

j sia HUMAN OVA FROM LARGE FOLLICLES

TABLE 1-Continued

15

DAY OF | ;UMBER . CONDITION OUST | soe | Ea eg] welts | CUPSTE pean Mm. 8P 26 10th =| L.Ov.,7 1==10 All ova degen-\Ovary also con2== 6 erating tained 1 large 38= 5 corpus 1== 3 9P 26 10th | 10v.,19|1—13X12X10 One ovum very2 ova in largest degenerate follicle 10P | 37 | llth | 10v.,7 | 1=13X12x9 |All atretic, sev2== 8 eral in very 1= 7 early stages 1== 6 2== 2 8A lith 1Ov.,7 | 1=28X18 X16 |Largest con-|\No recent cor5=8 to 6 tained normal] pora in this ovum ovary 31A 12th 1 Ov., 4 | 1=20X 20X16 |Largest follicdleOva in 5-mm. 8= 5 normal follicles float. ing free 16B 12th 1 Ov., 2 1= 8 Larger follicle|Positive hor1= 5 normal mone test (A no. 31) 11P 44 12th 1 Ov., 2 1= 6 Larger normal,|(No, 100) l= 5 smaller atretic| 12P | 28 12th 1 Ov., 7 1=10 10-mm. follicle|Mitoses (no. normal (102). In eumulus 11A 13th 1 Ov.,7 | 1==24% 22X19 [Largest follicle\(Other ova not 1=11 atretic sectioned l== 9 4== 7 20A 13th 1 Ov., 4 | 1=27X 24X18 |Largest atretic;|One of 7-mm. 3= 7 follicles _nor| mal (no. 0-25) 25A 13th | Part Ov. | 1=25x 2220 [Atretic 4.2 ee. liquor folliculi 23B 13th 1 Ov. 1=22, Atretic Cumulus gone several smaller 13P 26 13th or | 10v,,6 | 1=15X12X12 ,|Two largest fol--Other ova not 14th 1=16 licles atretic | sectioned 1==10 1== 8 2Q== 5

THE AMERICAN JOURNAL OF ANATOMY, VOL. 46,

no. 1 16

EDGAR ALLEN ET AL.

TABLE 1—Continued

DAY OF NUMBER case | son | Bx, | Nor | ySiBROR, | SoNDITION | eaanns NO. AY |roLLrcLEs 2 ifm. : 4A 14th 1 Ov., 6 2=11 Two larger fol= § lieles atretic 18A | 45 14th | 10v.,9 1= 9 All atretie, 3= 8 largest. hemor3= 7 rhagic 2= 5 28A 14th 1 Ov., 6 1= 9 2 of the larger|3 large corpora 2= 8 atretic, 2 nor-| lutea in this 1=7 mal ovary a= 5 5B 14th 1 Ov., 5 1= 8 1 of 6-mm. fol-/Others not sec4= 6 licles atretic | tioned 14P | 40 | 14th 1 Ov., 2 1=10 Both normal |Largest con1= 6 tained mature ovum (no. 117), smaller growing 15P | 33 14th 1 Ov., 6 1=11 4 of these were/Tubal ovum V 3= 8 atretic from this ovary; recent corpus 6A 15th 1 Ov., 2 Larger atretic|/Recent corpus smaller grow in this ovary ing 15A 15th 1 Ov., 4 | 1=16 1613 1 large corpus 2=>7 in this ovary 1= 6 23A 15th | Part of l= 9 Atretic? Tubal ovum I 1ov., 3 1= 6 | from this 1= 3 | ovary; recent

| corpus 12A 16th | 1 Ov.,11 6=8tol0 |One 9-mm. fol41 recent corpus 5=3 to 6 licle normal | luteum 16A 16th | 1 Ov.,9 6=6 to 7 |One 7-mm. fol{Others not see: 3= 4 licle atretie | tioned 19A 16th 1 Ov., 1 1=10.5 ‘Ovaries _sclerotic 16P | 33 ‘ith |i 1 Ov.,4 1= 6 Largest atretic|Others not sécSS 2= 5 | tioned I= 2 ! HUMAN OVA FROM LARGE FOLLICLES

TABLE 1—Concluded

17

CASE uen- | NUMBER SIZE OF conprtion | no. | AGE eTRUAL | FOLLICLES! FOLLICLES | OF OVA | REMARKS | ° Mm. 22A 20th 1 Ov.,5 = 9 ‘All atretic jl large corpus B= 5 | luteum 1= 4 7B 21st L.Ov., 4 2=5to 9 Hemorrhagic bo 2=4to 5 | | R. Ov., 5 3=7to 9 |Hemorrhagie | : 2= 6 214 | | 22nd | Part of : 1= 6 » 1Ov.,1 | Q7A | | 23rd: 1 Ov. 4 | 1=13X11X10 ‘Ovary also coni 1= 9 tained _— large | l= 7 hemorrhagic | . 1=4to 5; | cyst 9B 25th | L.Ov., 4 |. | | R.Ov., 6 | 17P 30 | 4 weeks , 1 Ov.,8 1= 8 All atretic, 4 in | 1=7 | very early 2= 6 stages | | 38= 5 1= 4 18P |; 40 | 1 Ov., 11! 2= 7 One 4-mm. fol-/Menstrual _his| 6= 4 Niele normal,| tory uncertain

3==2to 3 | others atretic” 19P| 31 | Eetopie | 1 Ov.,5 1= 8 4 larger were in " preg. 38= 6. early atresia, ard l= 4 smallest normonth | mal “oP | 37 Full-term: 1 Ov., 4 l= 4 Largest atretic Ovary containpreg.; | 2= 3 ed full-term caesarian | l= 2 corpus luteum 5A Full- | 1 Ov., 14 2.5— 4 Ova not see term tioned ‘preg. | 9A Ful | L.Ov.,,8| 6=4.5to6 Ovary - containterm R.Ov., 3 == 8 ed full-term preg. 3=4 to 6 corpus luteum 18 EDGAR ALLEN ET AL.

Even though the ovum may have become detached from the follicle wall and be floating free, there are usually very few detached follicle cells to be found in the liquor.

The granulosa layers of the follicle wall vary between four and ten cells in thickness. In representative sections of walls of large follicles, where the presence of cells in mitotic division has shown the follicle in question to be normal and actively proliferating, at least four layers of follicle cells could be counted.

The living human ovarian ova

During the course of this work we have been able to observe the following points concerning the living ova examined in liquor folliculi immediately after removal of the ovary at operation. These observations confirm in many details earlier ones made by Nagel and others. The human ovum is very transparent. The cytoplasm is usually almost colorless or a light yellow and finely granular in appearance under high magnification. If the egg is not surrounded by too many layers of follicle cells, the nucleus is often visible and occasionally a nucleolus may be observed. .

In ova with nuclei intact which on section proved to be normal, the vitellus lay directly against the zona pellucida without the intervention of an appreciable perivitelline space. Nagel described the presence of a perivitelline space which might permit rotation of the ovum within the zona pellucida in living human eggs. Such rotation would allow for orientation to gravity as observed in the eggs of some other animals. Since a large majority of oocytes recovered from large follicles were shown after sectioning to have begun degeneration, it seems doubtful whether an appreciable perivitelline space appearing before the beginning of maturation can be considered a normal attribute or a degeneration artefact. A later cytological description of normal oocytes shows little or no perivitelline space before the beginning of maturation.

In the course of early work on this problem a beautifully developed corona radiata was observed about a living ovum HUMAN OVA FROM LARGE FOLLICLES 19

from a large follicle obtained from a cow in oestrus. So far in this study of hnman ova we have not obtained a specimen with a fully developed corona. Several ova have been recovered, however, in which a beginning of corona formation was indicated. Further search should locate better specimens.

Earlier reports (Nagel) of amoeboid motion in nucleoli of living ova are interesting. It seems probable, however, that observations of ovarian ova which were sufficiently free from surrounding follicle cells to allow clear distinction of the nucleoli could hardly be considered normal.

Measurements were made of eighteen fresh ova from large follicles which (after section) proved to be normal or in very early stages of atresia. Measurements were made as soon as possible after recovery while the ova were floating in liquor folliculi. It is necessary that the surrounding fluid be deep enough to support the ovum, for otherwise the fresh specimens may flatten out and considerably increase the horizontal diameters. In some cases, due to surrounding follicle cells, outlines of the vitellus could not be clearly enough defined for accurate measurements, and therefore only outside diameters including the zona pellucida were taken. The ova measured were taken from follicles ranging between 7 and 10.5 mm. in diameter, with the exception of one 20-mm. follicle. The range of measurements of greatest outside diameters of eighteen ova was from 117 to 142u (average, 123 y). The vitellus ranged in size from 84 to 108 » (average, 95 4) (six ova). The zona pellucida varied from 12 to 18 u in thickness (five ova). A comparison of these measurements with others taken after the ova were sectioned shows a shrinkage of at least 10 to 15 per cent due to fixation and dehydration. The zona pellucida after fixation is usually considerably thinned, Measurements of human tubal ova are available in table 2 of the writers’ earlier paper for comparison with those of ovarian ova. Reference should also be made to Hartman’s paper (’29) on the size of mammalian ova.

It is interesting to note from examination of the ovaries recovered on the second and fifth days of the menstrual cycle 20 EDGAR ALLEN ET AL.

(eases 1, 2, and 3P, table 1) that, although many follicles larger than 5 mm. were present, all ova examined were undergoing degeneration. None of these follicles had a chance of being the one which might have been favored by normal ovulation in the next seven to twelve days. On the seventh day of the menstrual cycle an ovary obtained from case 4P contained one normal 7-mm. follicle—others larger than 4 mm. were atretic. On the eighth day of the cycle one of five large follicles contained a normal ovum. As will be noted by a study of table 1, a large majority of all of the follicles visible on the surface of the ovary contained ova which had begun to degenerate. Take, for instance, case 6P, an ovary removed on the ninth day of the menstrual cycle. This ovary contained seventeen large follicles, of which only four were normal. These follicles could not be distinguished from the others on gross examination. The largest one, 16 < 12 * 11 mm. in diameter, contained an ovum still surrounded by a dense cumulus which attached it to the wall of the follicle. The cells of the cumulus were normal, although no mitoses could be identified with certainty; consequently, it is doubtful if this follicle was still growing. There were no signs of a corona radiata in the cells of this cumulus. The zona pellucida and vitellus were both normal. The nucleus was slightly off center, the nucleolus distinct. This ovum measured after fixation 100 & 93 in the central section. The vitellus alone measured 89 & 83. The nucleus was 23 » in diameter after fixation. Three other follicles (one 7, one 5, and one 4 mm.) were normal. All others showed definite signs of atresia. It is probable therefore that only one of the four normal follicles could have been ovulated in course of the next five days and that the other sixteen would probably have been eliminated before the appearance of the next menses nineteen or twenty days hence. It must be recalled that only one ovary from this case was examined and that there is no way of telling whether the ovum due to be ovulated would have come from this or the opposite ovary. Consequently, in this calculation the number of follicles elimHUMAN OVA FROM LARGE FOLLICLES 21

inated should be doubled. Data of this sort for other cases will be found in column 6 of table 1.

Description of normal ova sectioned after removal from large follicles

Ovum 5P (fig. 1) was taken from a 4-mm. follicle in an ovary removed from a case of ectopic pregnancy of about the third month (case 19P, table 1). In the same ovary there were four other follicles, three of 6 mm. and one of 8 mm. in diameter—all atretic. Ovum 5P was surrounded by a dense cumulus which attached it to the follicle wall. Many cells of the cumulus were in mitotic division, indicating rapid growth of this follicle. The nucleus was centrally located, with a prominent nucleolus and masses of chromatin. The cytoplasm was uniformly granular.

Ovum 23B was also obtained from a 4-mm. follicle, but from an ovary removed on the thirteenth day of the menstrual cycle. This ovary contained a recent corpus luteum and also a degenerating cystic follicle 22 mm. in average diameter. Ovum 23B was surrounded by a dense cumulus still attached to the follicle wall. The nucleus was slightly off center and contained a prominent nucleolus. The ovum was cut into twelve sections. The diameters of the sixth section were 112 & 85. The zona pellucida was 5 to 6 thick.

Ovum 49P (fig. 2) was removed from a 5-mm. follicle from an ovary which contained sixteen other follicles (case 6P, table 1), several of which were normal. Its densely packed cumulus still contained mitoses. The nucleus was slightly eccentric, the nucleolus and several chromatin masses were normal. Several interesting cytoplasmic inclusions are shown in the photograph.

Ovum 112P was from a 6-mm. follicle removed four weeks after the last menses (case 17P). The nucleus was centrally located and contained a clearly defined nucleolus. The vitellus was normal and closely applied to the zona pellucida without any intervening perivitelline space. Ovum 110P was from another 6-mm. follicle from the same ovary. This ovum was 22 EDGAR ALLEN ET AL,

surrounded by a dense cumulus. It had a thick zona pet lucida tightly applied to the vitellus.

Ovum 118P was also from a 6-mm. follicle from an ovary removed on the fourteenth day of the menstrual cycle (case 14P, table 1). This ovary also contained a rapidly growing 10-mm. follicle from which an ovum in maturation divisions was obtained. Ovum 118P was still attached to the follicle wall by a dense cumulus, in which at least ten mitotic figures were counted. There was a slight development of a corona radiata about this ovum. The nucleus was peripherally located.

Ovum 42P was from a 7-mm. follicle from an ovary of a thirty-two-year-old woman removed the ninth day of the menstrual cycle (case 6P, table 1). There were sixteen other follicles in this ovary. Ovum 42P (fig. 3) was still attached to the wall of the follicle, but the cumulus contained several lakes of secondary liquor folliculi. In the two or three layers of follicle cells immediately surrounding the ovum a beginning of corona-radiata formation was apparent. The zona pellucida was thick and darkly stained (iron alum hematoxylin usually stains the zona deeply) and closely applied to the vitellus. The vitellus was uniformly granular in texture. The nucleus was spherical and peripherally located. It contained a large nucleolus (in an adjacent section which was not photographed) and four or five small chromatin masses. That this ovum was normal and the follicle still actively growing is shown by the presence of mitoses in the cumulus. Two of these are shown in figure 4.

Ovum 73P was from a 10-mm. follicle removed on the ninth or tenth day of the menstrual cycle (case 7P) from an ovary which contained thirty large follicles. A persistent cumulus attached the ovum to the wall of the follicle. The nucleus was peripherally located. No mitoses could be identified with certainty; consequently, it is doubtful if this follicle was still growing.

Ovum 102P (fig. 5) was from a 10-mm. follicle removed on the twelfth day of the menstrual cycle (case 12P). There was HUMAN OVA FROM LARGE FOLLICLES 93

a slight indication of a corona radiata about this ovum. The cumulus contained several lakes of secondary liquor folliculi. At least eight cells in mitosis were counted in the cumulus. The nucleus was centrally located and contained a nucleolus and four or five chromatin masses. (The section photographed included only a thin slice from one side of the nucleus.) The vitellus was finely granular in appearance and there was no appreciable perivitelline space.

Ovum 35A (fig. 6) was from a 9-mm. follicle in an ovary removed on the twenty-third day of the menstrual cycle (case 22A). A dense cumulus attached it to the follicle wall. A total of thirty cells in mitosis were counted in twenty-five sections of the cumulus. Two are shown at opposite sides of the ovum in figure 6, one at the upper right, the other at the extreme lower left corner of the figure. Both are in the anaphase of division, the one to the right being adjacent to the zona pellucida, the one to the left being six to eight cell layers distant. There was a slight indication of corona formation around this ovum. The plane of the section photographed cut through one side of the nucleus, therefore nuclear details are not illustrated.

Another similar follicle was present in this same ovary, both were growing rapidly. Since this ovary was removed on the twenty-third day of the cycle, the next ovulation would not have been due for nearly three weeks. There seems a reasonable doubt as to whether ova from either of these two follicles might have survived this interval and been ovulated.

Ovum 125P (fig. 7) was recovered from a 12-mm. follicle. There were two or three layers of adjacent follicle cells in which a corona radiata had begun to form. The nucleus was centrally located and contained a clearly defined nucleolus and several chromatin masses. This egg was cut into eleven sections. The sixth section measured 96 X 91 u.

Ovum 126P (fig. 8) was from a 10-mm. follicle from the same ovary as 125P. It was also attached to the follicle wall by a cumulus containing many mitotic figures, one of which is shown in the upper center of the photograph. The nucleus was slightly off center. 24 EDGAR ALLEN ET AL.

Ovum 34A was from a folliele 138 * 11 * 10 mm. removed on the twenty-third day of the menstrual cycle (case 27A). This ovum, although surrounded by a dense cumulus, was floating free in the liquor folliculi. Formalin fixation collapsed the zona pellucida. A careful count of mitotic figures in the cumulus showed a total of fifty-one, indicating extreme hyperplasia. The vitellus had a narrow zone of clear exoplasm. The nucleus was peripherally located and contained a nucleolus and several masses of chromatin. From the above description and the time in the cycle it is doubtful if this ovum could have been normally ovulated.

Ovum 38A (fig. 9) was obtained from a follicle 20 x 20 « 16 mm. in diameter removed on the twelfth day of the menstrual cycle (case 831A). There had been no recent ovulation in this case. The follicle was so large that on gross examination it was thought to be cystic. It was surrounded by a dense cumulus in which several small lakes of secondary liquor folliculi had formed. There was a suggestion of corona formation. Several mitoses could be definitely identified among the follicle cells. The nucleus was centrally located and contained a nucleolus with a clear center and dark circumference. This is the largest normal follicle which was obtained in this study. As it was larger than most human

corpora lutea, it was probably an unusually large normal follicle.

Maturation division of the human ovum

The main objective of this investigation, a search for oocytes in maturation stages, has resulted in the finding of one ovum (117P from case 14P) which had reached the stage at which ovulation occurs in most mammals. This oocyte (figs. 10, 11, and 12) was recovered from a 10-mm. follicle on the fourteenth day of the menstrual cycle. The ovary containing this follicle was removed from a patient forty years of age (case 14P, table 1). It also contained a growing 6-mm. follicle already described (ovum 118P). HUMAN OVA FROM LARGE FOLLICLES 25

This oocyte contained a peripherally located group of chromosomes representing the second maturation spindle and also a first polar body. It was sectioned at a thickness of 10 and cut into twelve sections. The other dimensions of the sixth section were 85 X 76, including the zona pellucida, and 73X65, for the vitellus alone. These measurements represent the size after fixation and should not be confused with measurements of fresh ova, because fixation causes considerable shrinkage. The median section of this ovum cut through the group of chromosomes of the second maturation spindle, some chromosomes being located in each of sections 6 and 7 (located at 11 o’clock in figures 11 and 12). A total of twenty-two chromosomes could be counted, with the possibility that one or two of these were double. This number closely approximates the haploid count of twenty-four obtained by Painter (’23) in his study of spermatogenesis in man. Counts of eighteen, fourteen or fifteen, and eighteen were obtained in human tubal ova II, ITI, and V, respectively.

It was not possible to distinguish fibers of a second meiotic spindle, due perhaps to a partial polar view of the chromosomes in an equatorial plate. Nor was it possible to identify chromosomes as dyads.

The first polar body was located in the tenth section (fig. 10). It was possible to identify with certainty twelve or thirteen chromosomes in this polar body.

The zona pellucida was intact in all places and 4u thick after fixation. The vitellus was of a uniformly granular appearance without differentiation of a lighter exoplasmic zone. <A definite vitelline membrane and narrow perivitelline space were present.

Parts of the cumulus consisting of rather densely packed cells were still in place on one side of the ovum, while on other sides these cells had been dispersed. Part of the dispersion was undoubtedly due to difficulties in fixation, dehydration, and embedding of the ovum. This point should be kept in mind when comparing photographs of the ova 26 EDGAR ALLEN ET AL.

described in this paper with ova sectioned in intact follicles. Removal of the ovum from the follicle and subsequent fixation and dehydration often disperse much of the cumulus. There were no signs of a corona radiata. In the twelfth section, which contained only a small part of the ovum, three or four cells were located which were undergoing mitosis. In other regions a few pycnotic cells were present. From a study of the parts of the cumulus which are still present it is probable that this oocyte was a normal second meiotic division and that the follicle was still actively proliferating. Since it was obtained on the fourteenth day of the menstrual cycle and there had been no recent ovulation, it is possible that this oocyte might have been normally ovulated. As far as the writers are aware, this is the first human oocyte to be recovered while in the maturation process from a follicle which might normally have ovulated.

Pre-ovulation size of the normal follicle in the human ovary

This finding of one ovum in maturation division in a normal 10-mm. follicle indicates that at least 10 mm. might be considered pre-ovulation size for a human follicle. We have several cases, however, in which follicles having a greatest diameter of 18 and 22 mm. (follicles which were tentatively diagnosed as probably cystic on gross examination) were found to contain normal ova and proliferating follicle cells. Where these cases were found in intermenstrual ovaries, it is quite possible that the ova might have been normally ovulated. It is doubtful, however, if the normal pre-ovulation size of a follicle often exceeds the size of the fully developed corpus luteum.. An average size of the human corpus during the third week of the menstrual cycle, computed from data obtained in this work, would range between 15 and 18 mm. A follicle from 10 to 15 mm. in diameter during the early intermenstrum might therefore be considered normal, and even larger follicles, if their position in the ovary caused flattening, might also be of normal pre-ovulation size. HUMAN OVA FROM LARGE FOLLICLES 27

Relation of the point of attachment of the ovum to the follicle wall to the probable site of the future rupture point

Recently, Thomson (’20), Strassmann (’23), Shaw (’25), and others have stressed the importance of the point of attachment of the ovum to the site of the probable rupture point. Thomson believes that the human granulosa layer is stripped off and expelled with the ovum at ovulation. Shaw accepts the theory proposed by Strassmann that if the point of attachment of the ovum to the follicle wall is toward the center of the ovary instead of toward the surface, the granulosa layers rotate within the theca as the follicle works its way to the surface preparatory to ovulation. Such an opinion hardly seems tenable after the splendid work on ovulation in ferrets (Robinson, ’18) and in rabbits (Hammond, 25). However, since the human material was available, it seemed worth while to make observations on these points.

Of the large follicles in which ova were still attached to the wall, the point of attachment in a majority of cases was not at the surface where the future point of rupture would be located. In dissecting large, follicles from the ovary the theca layers were often found to be very edematous. Even in these cases, however, there was sufficient attachment of connective tissues to prevent rotation of the granulosa layers within the theca. From this study of fresh ovarian tissues no evidence was found to support Strassmann’s theory of rotation of the granulosa layers within the theca before ovulation. The detachment of the ovum from the granulosa cells by the secretion of secondary liquor folliculi in intercellular spaces in the cumulus, as already demonstrated for several mammals, is shown in several specimens of human ova already described. There seems, therefore, little necessity for such a theory as that proposed by Strassmann to explain the liberation of the ovum. 28 EDGAR ALLEN ET AL.

Atresia in large follicles

A sure sign of atresia to be noted as a follicle is dissected open is cloudy or gelatinous liquor folliculi. Occasionally, small masses of dispersed follicle cells may be found in liquor from normal follicles, but usually such specimens show early signs of atresia. In all cases where the inner surface of the follicle wall showed projections of the granulosa similar to villi, the follicle proved to be atretic. Sometimes small, hollow spheres of cells (‘casts’) were found floating free in liquor folliculi of atretic follicles. The complete dispersion of the cumulus or the occurrence of considerable numbers of degenerating follicle cells in an intact cumulus was often found (figs. 19, 20, and 23). These must be considered definite signs of atresia of the follicle, for the human ovum, like that of several other mammals (dog, rabbit, monkey, and others), still retains several layers of normal follicle cells during ovulation (figs. 11, 12, and 16 of the writers’ paper on tubal ova).

Vacuolation of the yolk occurred in a considerable number of eggs (figs. 15 and 16). The vacuoles varied greatly in size, until in a few only a few shreds of reticulum remained (figs. 13 and 14). In a few ova the vitellus had definitely fragmented. Another degenerative process apparent in the yolk was the appearance of blotchy masses interspersed with clearer globules of cytoplasm. A still earlier stage of cytoplasmic degeneration was the appearance of a clear outer zone and a dark-staining central portion of the vitellus (figs. 20 and 26). Figure 19 shows the position of these zones reversed.

The most evident signs of atresia in the nucleus were shrinking or fading of nuclear structures (fig. 21), crinkling of nuclear membranes (fig. 28), and disappearance of nucleoli. Also the migration of a vesicular nucleus from a central to a peripheral location (figs. 17, 21, and 28), usually regarded as one of the first preliminaries of maturation divisions, might be considered a sign of early atresia if obtained after HUMAN OVA FROM LARGE FOLLICLES 29

the fifteenth day of the menstrual cycle, for such an ovum would have little chance of existing for a month until the next ovulation.

One type of atresia which was observed in several progressive stages consisted of an invasion of follicle cells through the zona pellucida (figs. 25, 26, 27, 29, and 30). In one of the most striking cases more than one hundred follicle cells had entered the perivitelline space, and the outer zone of cytoplasm was quite clear and finely granular, while the central portion was coarsely granular (fig. 26). This process of follicle-cell invasion may occur in the tubes as well as in follicles, as noted in human tubal ovum V, where the yolk was quite badly fragmented and a considerable number of follicle cells had penetrated the zona pellucida. A still more extreme stage of this same process was found in a large follicle where a persistent zona pellucida enclosed more than one hundred follicle cells and it was difficult to distinguish any other remains of the ovum (fig. 30).

Degenerating ova in large atretic follicles

One of the most interesting ovaries from the point of view of atretic follicles was that from which tubal ovum V had been ovulated (case 15P, table 1). It was removed on the fourteenth day of the menstrual cycle and contained, beside the recent corpus luteum corresponding to tubal ovum V, six large follicles, of which one was 11 mm. and three 8 mm. in average diameter. The 11-mm. follicle contained a misshapen ovum (121P) from which the nucleus had disappeared. The surrounding follicle cells had been considerably dispersed and formed a loose gelatinous mass about the ovum. Ovum 122P, from an 8-mm. follicle, also was surrounded by a partly dispersed cumulus. The nuclear membrane was crinkled and the chromatin massed. Ovum 123P, also from an 8-mm. follicle, was similar, but in a slightly more advanced stage of atresia. An ovum from another 8-mm. follicle from this ovary was lost during the embedding process. Ova from the other two follicles were not sectioned. It is interesting to 30 EDGAR ALLEN ET AL.

compare these ova with tubal ovum V, which was also quite degenerate. One wonders if these follicles and that containing tubal ovum V started to develop on an equal footing at some fairly definite period earlier in the cycle and what may have determined that tubal ovum V rather than one of these should attain ovulation.

Ova 115P and 116P (figs. 13 and 14) were twins from the same 10-mm. follicle removed from a thirty-six-year-old patient. A definite menstrual history of this case was not obtained. Each of these ova had a separate cumulus composed of densely packed, pyenotic, dark-staining follicle cells. In each ovum the zona pellucida was intact, but all that was left of a vitellus was a suggestion of a coarse reticulum. Figure 15 shows an earlier stage of this same sort of atresia.

Ovum 76P was from a 3-mm. follicle from an ovary of case 7P. It was still surrounded by a considerable cumulus which contained several small lakes of fluid each surrounded by a single cuboidal or low columnar cell layer. These formations resembled the ‘bodies of Call and Exner,’ and it seems probable that when further isolated by dispersion of the cumulus they might give rise to the ‘molds’ or ‘casts’ of cells sometimes found in atretic follicles. So far these ‘casts’ have not been observed in normal follicles.

Atretic ovum 27P was from a 7-mm. follicle in an ovary removed on the fifth day of the cycle (case 3P, table 1). The cumulus was dispersed atid many of the follicle cells were pycnotic. Observed before fixation, they looked like miniature frog’s eggs in a gelatinous mass. The zona pellucida was frayed and the cytoplasm had a moth-eaten appearance. The nuclear membrane was indistinct and the ‘chromatic material was distributed in large blotches.

Ovum 40P (fig. 16) was from a 4-mm. follicle removed on the ninth day of the cycle (case 6P). It was floating free in a partly dispersed cumulus. The zona pellucida was frayed, the cytoplasm shrunken and vacuolated, and the nuclear membrane crinkled. A large globular nucleolus was clearly defined. HUMAN OVA FROM LARGE FOLLICLES 31

Ovum 15P (fig. 17), from a 6-mm. follicle from an ovary removed on the eighth day of the cycle (case 5P), was partly surrounded by a dispersed cumulus many cells of which were pycnotic. A large vesicular nucleus with a prominent nucleolus was located slightly off center. Ova 16P and 17P (figs. 18 and 19) were from 7- and 6-mm. follicles, respectively, from the same ovary. There were two other large atretic follicles in this ovary. The zona pellucida of ovum 16P was ragged. Ovum 17P (fig. 19) showed a more advanced stage of a similar sort of atresia. In addition the nuclear membrane was crinkled and the cytoplasm was blotchy with the exception of a clear perinuclear zone.

Ovum 2P (fig. 20) was a degenerating ovum from a 6-mm. follicle from an ectopic pregnancy of three months’ standing (case 19P). The cumulus had been almost completely dispersed. The nucleus was intact, but peripherally located. A nucleolus and several masses of chromatin appeared quite normal. The central mass of cytoplasm was coarsely granular, while the outer zone was clear.

Ovum 37A (fig. 21), from a 4-to-5-mm. follicle, appeared fairly normal, but it was floating free and its cumulus was flattened and contained many pycnotic cells. Nuclear structures were especially clear. This ovary was from case 27A and was removed on the twenty-third day of the cycle. Ovum 36A was from a 7-mm. follicle from the same ovary. The nucleus appeared to be quite normal, but the vitellus was shrunken and vacuolated.

Ovum 28A was from a 9-mm. follicle removed on the twentieth day of the cycle (case 22A). This was a binucleate ovum. A nucleolus could be identified in each nucleus. The ovum was normal except for vacuoles in the vitellus. The cumulus was quite dense and contained several cells in mitosis, indicating that the follicle was still growing. Ovum 29A, from a 5-mm. follicle of the same ovary, had an eccentric nucleus. Also the cumulus contained many degenerating cells, although it still attached the ovum to the follicle wall.

THE AMERICAN JOURNAL OF ANATOMY, VOL. 46, NO. 1 32 EDGAR ALLEN ET AL.

Case 20A was very interesting and consequently merits description in greater detail. The patient was operated on the thirteenth day of the cycle. There were no corpora lutea present in either ovary and yet there was a history of regular menses for the last several months. This was a clear case of several menstrual periods without ovulation. The ovary removed contained four large follicles. The largest (27 x 24 X18 mm.). was probably cystic. Two and eighttenths cubic centimeters of clear fluid were recovered. The collapsed wall of the follicle weighed 0.9 gram. Both fiuid and follicle wall contained considerable amounts of hormone. (These results may be found in the writers’ paper on the hormone content of ovarian tissues.) The follicle wall was lined with five or more layers of loosely packed granulosa cells. No mitoses were found. In some regions infiltrations of leucocytes were present. The ovum recovered from this follicle still had six or seven isolated follicle cells adherent to it. The ovum was almost spherical (67 »p in diameter), except for a bulge at one point. The greatest diameter through this bulge was 80. There was no zona pellucida apparent, the few follicle cells appeared as though attached directly to the surface of the vitellus. This ovum fragmented during dehydration, and consequently a study in section was not possible. The follicle with its contained ovum was undoubtedly atretic.

There were also three 7-mm. follicles in this same ovary. In two of the three the ova were attached by dense cumuli to the deep walls of the follicles. These follicles were normal.

Ova 18A and 21A were obtained from 7-mm. follicles dissected from an ovary removed on the fourteenth day of the cycle (case 18A). They were very similar and might have been considered normal except for the dispersed cumuli and eccentric nuclei. Both had clear-cut, red-staining nucleoli with closely massed chromatin.

Ovum 4A (fig. 22) was recovered from a follicle of 28 x 18 X 16 mm. from an ovary removed on the eleventh day of the menstrual cycle (case 8A). Also 1.8 cc. of follicular fluid obtained from this specimen returned posiHUMAN OVA FROM LARGE FOLLICLES 33

tive tests for the presence of hormone. A section of the follicle wall (fig. 23) showed six or seven layers of granulosa cells, with a columnar layer bordering the theca interna and an occasional mitotic figure. The ovum (fig. 22) was somewhat distorted, but still surrounded by two or three layers of degenerating follicle cells among which partial development of a corona radiata was still evident. The nuclear membrane was broken down on one side, but the nucleolus was still prominent. This probably represents an early stage of atresia. There was also a 4-mm. follicle in advanced atresia in the same ovary.

Ovum 19B (fig. 28) was from a 22-mm. follicle removed on the thirteenth day of the cycle. It was surrounded by a large but partly dispersed cumulus containing a few pyenotic cells. The nucleus was peripherally located and the nuclear membrane crinkled. The photograph is of the third of twelve sections and consequently does not show the clear-cut edges of the zona and vitellus which are apparent in the equatorial plane, where they are cut at right angles.

Ovum 103P was from a follicle 138 * 12 * 10 mm. removed on the tenth day of the menstrual cycle (case 9P). The cumulus was almost completely dispersed, the vitellus vacuolated, and the nucleus shrunken and poorly staining. A second degenerating ovum was recovered from this same follicle, but was not successfully sectioned.

Ovum 24P (fig. 26) was from a 7-mm. follicle removed on the fifth day of the cycle (case 3P). The cumulus was partly dispersed and many of the follicle cells pyenotic. The zona pellucida was broken and more than one hundred follicle cells had invaded the perivitelline space. The vitellus had an outer clear zone and a central zone of coarsely granular reticulated appearance. In this respect this ovum was similar to that shown in figure 20. The nucleus was spherical and centrally located. The nucleolus was distinct.

Ovum 84P (fig. 25) was another case of extreme atresia. It was obtained from a 4-mm. follicle removed from the ovary of a forty-two-year-old patient (case 18P). The ovum was 34. EDGAR ALLEN ET AL.

very small, being cut into only three sections at a thickness of 10yu. The zona pellucida was still distinct, but shrunken and misshapen. Several follicle cells had invaded the perivitelline space. There remained only a thin cloud of cytoplasm. Ovum 88P was from another 4-mm. follicle from the same ovary. The cumulus was almost completely dispersed, the outer surface of the zona ragged, the nucleus poorly stained, with just a trace of nuclear membrane and no visible chromatin material.

Ovum 108P (fig. 27) was from a 5-mm. follicle removed during the fourth week after the onset of the previous menses (case 17P). The ovum was very atretic. The zona pellucida was thin and extremely enlarged, increasing the perivitelline space, which was crowded with follicle cells. The vitellus was intact and finely granular in appearance. The nuclear membrane had disappeared, but several inclusions in the cytoplasm might. possibly represent the remains of nucleolus and chromatin. It is not difficult to visualize a continuation of atresia of such a specimen, leaving only a ‘cell ball’ or ‘bag of cells’ (zona pellucida filled with follicle cells) similar to specimen 127P (fig. 30) or some of the specimens recovered from the tubes.

Ovum 25P (fig. 29) was a more advanced stage of the same type of atresia. It was recovered from a 4-mm. follicle on the fifth day of the menstrual cycle. Only part of the vitellus was recognizable, enclosed with many follicle cells in a broken zona pellucida. Ovum 91P, from a similar follicle of this same ovary, had no nucleus and was almost completely free from follicle cells.

Ovum 22P was from an 8-mm. follicle in advanced atresia. The ovum was just a shadow. The zona was ragged and had only a few degenerate follicle cells still adherent to it. The nucleus and nucleolus were still recognizable.

Ovum 51P was from a 3-mm. follicle removed on the tenth day of the cycle (case 8P). The perivitelline space contained fourteen or fifteen follicle cells which had invaded the zona pellucida. The ovum in other respects seemed fairly normal. HUMAN OVA FROM LARGE FOLLICLES 35

Ovum 54P was from a 5-mm. follicle removed on the second day of the cycle (case 1P). The nucleus was recognizable, but the vitellus was shrunken and vacuolated.

Ovum 127P (fig. 30) was very degenerate. It was obtained from a 10-mm. follicle from the ovary of a twenty-four-yearold patient on the seventeenth day of the menstrual cyele. All that remained was a group of cells (probably follicle cells) enclosed in a degenerate zona pellucida. This specimen, representing an advanced stage of follicle-cell invasion into atretie ova, is quite similar to one type of ‘cell ball’ found in the uterine tubes during our search for tubal ova. The intermediate stages of this process shown in plate 5 leave little doubt as to the follicular origin of some of these ‘cell balls.’

DISCUSSION

The very small percentage of normal follicles among those visible on the surface of the ovary is noteworthy. This is to be expected, however, when one realizes that usually only one follicle from both ovaries is to be ovulated once a month. Still the usual opinion, unless considerable thought has been given to the matter, seems to be that some of the smaller visible follicles are to be held in a condition of suspended metabolism, to ripen during later cycles. From the present study it seems probable that, of the many follicles large enough to be visible on the surface of the ovaries, none last until the following month. It is very probable that the one to achieve successful maturation and ovulation a month hence is at this time so small as not to be visible (or at least not prominent) on the ovarian surface.

These findings are of especial interest in relation to the hormone content of the human ovarian follicle. They place quite a different aspect on this question as concerns the human ovary than can be taken in regard to mammalian ovaries in animals where litters are produced. In the pig, where from nine to twelve large normal follicles can be expected in the two ovaries at oestrus, the hormone content 36 EDGAR ALLEN ET AL.

of liquor folliculi has been measured and found to be considerable (Allen and Doisy, ’27). It is quite probable that in the collection of liquor folliculi from human ovaries—a collection usually made without reference to normality of follicles from which the fluid is aspirated—the majority of material obtained must come from atretic follicles. This might easily explain the relatively low content of hormone obtained in some of these analyses.

The data obtained in this study support previous observations obtained in the study of pig ovaries (Allen, Kountz, and Francis, ’27) and indicate in human ovaries also that a considerable elimination of follicles occurs during late development. This is probably a selective elimination depending upon other factors, perhaps endocrine in nature, such as the anterior lobe of the hypophysis, as demonstrated by Smith and Engle (’27) and by Zondek and Aschheim (’27) for rodents.

The finding of only one ovum undergoing maturation divisions among more than 200 ova from large follicles that were sectioned, combined with Hinselmann’s failure to find such a stage in one hundred large follicles, emphasizes the difficulty of the search and the short duration of meiotic divisions. The specimens previously described by Dixon and by Hoadley and Simons were in very small atretic follicles. As far as the writers are aware, ovum 117P is the first human oocyte in meiotic division to be obtained at the ovulation time in the menstrual cycle and from a normal follicle which might have ovulated. At every possible opportunity search should be made for additional human oocytes undergoing maturation. This search should be focused on large follicles (9 to 15 mm. in diameter or even larger) between the twelfth and fifteenth days of the menstrual cycle. In this search the condition of the most recent corpus luteum should be noted. Since the presence of a newly forming corpus luteum indicates a recent ovulation, large follicles from such a case may reasonably be expected to be found in atresia. HUMAN OVA FROM LARGE FOLLICLES 37

It is very difficult to diagnose early atresia of large follicles, even after dissecting them from the ovary under the binocular microscope. Later atresia is evident in extreme edema of the theca layers. When the follicle is opened, if the ovum is floating free, it will probably be found to be degenerating, unless the investigator has stumbled upon a follicle in a pre-ovulation stage, for evidence from this work indicates that the normal human ovum remains attached by its cumulus to the follicle wall even in very large follicles until meiotic division begins. As in some of the lower mammals (see Robinson’s work on the ferret, ’17), it is probable that it is freed just before ovulation by secretion of secondary liquor folliculi dispersing the follicle cells at the base of the cumulus.

Since the primate ovum is usually surrounded by from two to four layers of follicle cells at the time of ovulation, dispersion of these adjacent layers of cells may be considered a criterion of early atresia. Mitoses may be present in these adjacent follicle cells up to the time of ovulation. Their absence might at least stimulate a search for other signs of early atresia. Pycnosis of follicle cells is a definite criterion of atresia.

Further advanced degeneration of ova in large follicles is evident in eccentric position of the nucleus (unless just preceding maturation), crinkling of the nuclear membrane or even disappearance of the nucleolus or the whole nucleus, vacuolation, fragmentation, or segregation of constituents of the cytoplasm. Another fairly common form of advanced atresia in large follicles is the invasion of cells (the majority probably follicle cells) under the zona pellucida. When this process is advanced, a ‘ball of cells’ enclosed in a degenerate zona pellucida may be all that remains of an ovum. Such specimens are sometimes encountered in washings from the uterine tubes. A closely graded series of stages from large follicles provides evidence for identification of some of these tubal specimens as degenerating ova. 38 EDGAR ALLEN ET AL.

SUMMARY AND CONCLUSIONS

1. Observations have been made on the number, size, and condition of the larger follicles in the human ovary at various stages of the menstrual cycle. Follicles of from 3 to 20 mm. average diameter were included. The number varied from two or three to thirty-two, average eight, per ovary. A great majority of these larger follicles were in the process’ of degeneration.

2. The method used was to open the follicle, recover the ovum, and section it separately. This permitted the utilization of follicular fluid from the same specimens for quantitative analyses of hormone content.

3. The presence of mitotic division of follicle cells of the cumulus or granulosa layers was used as a criterion of the normal growing follicle, for it has been shown in other mammals that proliferation of follicle cells continues up to the time of ovulation. Since it has also been demonstrated that the human ovum and that of the monkey may still be surrounded by two or three layers of follicle cells after ovulation, the persistence of at least this much of a cumulus about ova in large follicles seemed necessary for their being considered normal.

4. Descriptions of normal ova and follicles record the following additional major points: The follicle wall is usually lined with more than four layers of follicle cells. Normal ova were usually attached to the follicle walls, even in some of the largest follicles studied. The cumulus may be composed of densely packed cells or may contain small lakes of secondary liquor folliculi, the spread of which probably detaches the ovum before ovulation. Mitotie figures (as many as thirty and fifty in the cumulus of two specimens) are usually to be found in both cumulus and granulosa layers, The diameters of normal ova in large follicles varied from 117 to 142 (average, 123). The vitellus alone measured from 84 to 108 y (average, 95). The zona pellucida ranged from 12 to 18 y in thickness. These measurements were made of fresh ova in liquor folliculi. The vitellus is almost transHUMAN OVA FROM LARGE FOLLICLES 39

parent, with a slightly yellowish tinge, and until nearly the time of maturation is closely applied to the zona pellucida without the intervention of an appreciable perivitelline space. It is finely granular in texture. The nucleus is usually centrally located until the approach of maturation and contains a large nucleolus and several chromatin masses.

5. One oocyte was obtained which was in the process of meiotic division. This was recovered from a 10-mm. follicle on the fourteenth day following the onset of the previous menses. The chromosomes of the second maturation spindle were peripherally located. At least twenty-two were counted. Twelve chromosomes were counted in the first polar body. Mitotic division of follicle cells of the cumulus indicated that this follicle was normal and still actively growing. Since the oocyte was obtained at the proper time in the menstrual cycle, it is quite possible that it might have been ovulated.

6. Judging from the follicle described above and the several other larger follicles which contained normal ova, it seems probable that the pre-ovulation size of the human ovarian follicle may be at least 10 mm. in average diameter and perhaps even larger.

7. A great majority of the follicles studied were undergoing atresia. In many the cumuli had been partly or wholly dispersed and the ova were floating free in the liquor folliculi. Several stages of atresia are described. One interesting type was the invasion of follicle cells through the zona pellucida into the perivitelline space sometimes associated with vacuolation or fragmentation of the vitellus. This type of atresia may give rise to some of the ‘cell balls’ obtained from washing out the uterine tubes.

8. The data obtained in this study support the conclusion that radical selective elimination of ova oceurs in human ovaries during late stages of growth of the follicles.

THE AMERICAN JOURNAL OF ANATOMY, VOL, 46, NO, 1 40 EDGAR ALLEN ET AL.

BIBLIOGRAPHY

Appison, W. H. F. 1917 Fragmentation of the ovum within the Graafian follicles. Proc, Path. Soe. Phila., vol. 37, p. 1.

1927 The centenary of the discovery of the mammalian ovum. Med. Life, vol. 34, pp. 305-312.

ALLEN, Epgar 1922 The oestrous cycle in the mouse. Am. Jour. Anat., vol. 30, pp. 297-371.

——-—— 1927 The menstrual cye¢le of the monkey, Macacus rhesus: cbservations on normal animals, the effects of removal of the ovaries and the effcets of injections of ovarian and placental extracts into the spayed animals. Contributions to Embryology, no, 98, vol. 19, Carnegie Inst. of Washington, Pub. no. 380, pp. 1-44.

1928 An unfertilized tubal ovum from Maeacus rhesus. Anat. Ree., vol. 37, pp. 351-356.

1928 a Reactions of immature monkeys to injections of ovarian hormone. Jour. Morph. and Physiol., vol. 46, pp. 479-519.

ALLEN, Epear, AND Dorsy, E. A. 1927 Ovarian and placental hormones. Physiol. Reviews, vol. 7, pp. 600-650.

ALLEN, Epa@ar, Kountz, W. B., AND FRANCIS, B. F. 1925 Selective elimination of ova in the adult ovary. Am. Jour. Anat., vol. 34, pp. 455-467.

ALLEN, Epe@ar, Pratt, J. P., NEWELL, Q. U., AND BLAND, L. J. 1928 Recovery of human ova from the uterine tubes. Jour. Amer. Med. Assn., vol. 91, pp. 1018-1020. (See footnote 2, p. 2).

1929 Human tubal ova, ovaries, and uterine tubes. Anat, Ree., vol. 42, p. 2.

Aral, Hayato 1920 On the postnatal development of the ovary (albino rat), with especia] reference to the number of ova. Am. Jour. Anat., vol. 27, pp. 405-462.

Arnoutp, Luoyp 1912 Adult human ovaries with follicles containing several oocytes. Anat. Ree., vol. 6, pp. 413-422.

ASAMI, GOICHI 1920 Observations on the follicular atresia in the rabbit ovary. Anat. Ree., vol. 18, pp. 323-3844,

AvTHiAs, M. 1909 Les phénoménes de division de l’ovule dans les follicles de Graaf en voie d’atrésie chez le Lerot. Anat. Anz., Bd. 34, 8S. 1-23.

Barry, M. 1838 Researches in embryology. Phil. Trans. Roy. Soe. London, pt. 2, pp. 301-841.

1839 Researches in embryology. Phil. Trans. Roy. Soc. London, pt. 2, pp. 307-380.

Biscuorr, TH. L. W. 1845 Entwickelungsgeschichte des Hunde-Kies. Braunschweig.

Branca, A. 1925 L’ovoeyte atrésique et son involution. Arch. d. Biol., T. 35, pp. 325-440.

CuHaguTon, H. H. 1917 The fate of the unfertilized egg in the white mouse. Biol. Bull., vol. 33, pp. 321-331.

CuarK, E. B. 1923 Observations on the ova and ovaries of the guinea-pig, Cavia cobaya. Anat. Rec., vol. 25, pp. 313-338. HUMAN OVA FROM LARGE FOLLICLES 41

Corner, G. W. 1923 Ovulation and menstruation in Macacus rhesus. Contributions to Embryology, vol. 15, Carnegie Inst. Washington Pub. no. 332, pp. 75-101. 1928 Cytology of the ovum, ovary and fallopian tube. Special Cytology, Cowdry, vol. 2, pp. 1110-1150. Dixon, A. F. 1927 Human oocyte showing first polar body and metaphase stage in formation of second polar body. Irish Jour. Med. Sei., pp. 149-151. EncLe, E. T. 1927 A quantitative study of follieular atresia in the mouse. Am. Jour. Anat., vol. 39, pp. 187-203. Evans, H. M. 1916 On the behavior of the ovary and especially of the atretic follicles toward vital stains of the azo group. Anat. Rec., vol. 10, p- 264. FLEMMING, W. 1885 Ueber die Bildung von Richtungsfiguren in Siugethiereiern beim Untergang Graafscher Follikel. Arch, f. Anat. u. Phys., S. 221-244, Hammond, J. 1925 Reproduction in the rabbit. London, HaNnsEMANN, D. 1913 Uber den Kampf der Eier in den Ovarien. Archiy fiir Entw. der Organismen, Bd. 35. HartMaNn, CarL 1923 The oestrous cycle in the opossum. Am. Jour. Anat, vol. 33, pp. 353-395. 1926 Polynuclear ova and polyovular follicles in the opossum and other mammals, with special reference to the problem of fecundity. Am. Jour. Anat., vol. 37, pp. 1-52. 1929 How large is the mammalian egg? Quart. Rev. Biol., vol. 4, pp. 378-388. Heapr, W. 1900 The sexual season. Quart. Jour. Mic. Sce., vol. 44. Hents, J. 1873 Handbuch der systematischen Anatomie des Menschen. Bd. 2, Aufl. 2, S$. 504. Braunschweig. Hennecuy, L. F. 1894 Recherches sur |’atrésie des follicules de Graaf chez les mammiféres et quelques autres vértébrés. Jour. d. l’anat. et physiol., T. 30, pp. 1-39. Heyse, G. 1897 Einige Beitrage zur mikroskopischen Anatomie der Ovarien osteomalacischer. Arch. fiir Gynakologie, Bd. 53, S. 334. HINSELMANN, H. 1927 Maturity of the human ovum. Klin. Wochensehr., Bd. 6, S. 2289, Hoabuey, L., anp Simons, D. 1928 Maturation phases in human oocytes. Am, Jour. Anat., vol. 41, pp. 497-509. JOESSEL, G. 1899 Lehrbuch der topographisch-chirurgischen Anatomie. Zweiter Theil, fortgesetzt von W. Waldeyer, S. 793. Friedrich Cohen, Bonn. Keser, F., and Maui, F. P. 1910 Manual of human embryology. Kinerry, H. M. 1914 So-called parthenogenesis in the white mouse. Biol. Bull., val, 27, Kirguam, W. B. 1907 The maturation of the mouse egg. Biol. Bull., vol. 12, no. 4, pp. 259-265. KirkHam, W. B., anp Burr, H. §. 1913 The breeding habits, maturation of the eggs and ovulation of the albino rat. Am. Jour. Anat., vol. 15. 42 EDGAR ALLEN ET AL.

Lams, H., anp Doorme, J. 1908 Nouvelles recherches sur la maturation et la fecondation de l’oeuf des Mammiféres. Arch. de Biol. T. 23 (2), pp. 259-365.

Lors, L. 1911 The eyelic changes in the ovary of the guinea-pig. Jour, Morph., vol. 22, pp. 37-70.

1914 The correlation between the cyclic changes in the uterus and

the ovaries in the guinea-pig. Biol. Bull., vol. 27, pp. 1-44.

Lone, J. A., AND Evans, H. M. 1922 The oestrous cyele in the rat and its associated phenomena. Mem. Univ. California.

Lona, J. A., AND Marx, E. L. 1911 The maturation of the egg of the mouse. Carnegie Institute of Washington, Pub. no. 142.

LoneLtey, W. H. 1911 The maturation of the egg and ovulation in the domestic eat. Am. Jour. Anat., vol. 12, pp. 139-172.

McKenzigz, F. L. 1926 The normal oestrous cycle in the sow. Agricultural Experiment Station Bull. no. 86, pp. 1-41.

NaGeL, W. 1888 Das mensehliche Hi. Arch, f. mikr. Anat., Bd. 31, 8. 342-423,

Newman, H. H. 1913 Parthenogenetic cleavage of the armadillo ovum. Biol. Bull., vol. 25, pp. 54-78.

PaIntTER, TH. 1923 Studies in mammalian spermatogenesis. II. The spermatogenesis of man. Jour. Exp. Zodl., vol. 37, pp. 291-336.

Parkes, A. 8. 1926 Observations on the oestrous cycle of the albino mouse. Proce. Roy. Soe., B, vol. 100, pp. 151-170.

Pratt, J. P., ALLEN, Epgar, NEwELL, Q. U., AND BLAND, L. J. 1929 Human ova from the uterine tubes. Jour. Am. Med. Assn., vol. 93, pp. 834837.

Ropinson, A. 1918 The formation, rupture and closure of ovarian follicles in ferrets and ferret-polecat hybrids, and some associated phenomena. Trans. Roy. Soc. Edin., vol. 52, pp. 303-362.

RUBASCHKIN, W. 1905 Ueber die Reifungs- und Befruchtungsprocesse des Meerschweincheneies. Anat. Hefte, Bd. 29, 8. 507-554.

Sansom, G.S. 1920 Parthenogenesis in the water vole. Jour. Anat., vol. 55.

ScuréprEr, R. 1913 Ueber die zeitlichen Beziehungen der Ovulation und Menstruation. Arch. f. Gynak., Bd. 101, 8. 1-35.

1915 Anatomische Studien zur normalen pathologischen Physi ologie des Menstruationzyklus. Arch. f. Gynik., Bd. 104, 8. 27-102.

Sxitz, L. 1905-1906 Die Follikelatresie wahrend der Schwangerschaft. Arch. f. Gynak., Bd. 77, 8S, 203-356.

SHaw, W. 1925 Fate of graafian follicle in the human ovary. Jour, Obst. and Gynec. Brit. Emp., vol. 32, pp. 679-689.

SCHOTTLAENDER, J. 1891 Beitrag zur Kenntniss der Follikelatresie nebst einigen Bemerkungen iiber die unverdinderten Follikel in den Fierstécken der Sdugethiere. Arch. f. mikr. Anat., Bd. 37, 8. 192-238.

———-—- 1893 Ueber den Graaf’schen Follikel, seine Entstehung beim Menschen und seine Schicksale bei Mensch und Séugethieren. Arch. f. mikr. Anat., Bd. 41, 8. 219-294.

Situ, Smptima C. 1925 Degenerative changes in the unfertilized uterine eggs of the opossum, with remarks on the so-called parthenogenesis in mammals, Am. Jour, Anat., vol. 35, pp. 81-104, HUMAN OVA FROM LARGE FOLLICLES 43

SmiTH, PuILip E., AnD ENGLE, E. T. 1927 Experimental evidence regarding the réle of the anterior pituitary in the development and regulation of the genital system. Am. Jour. Anat., vol. 40, pp. 159-217.

Sosorra, J., AND BURCKHARD, G. 1911 Reifung und Befruchtung des Eies der weissen Ratte. Anat. Hefte, Bd. 42, 8. 433-498.

Stevens, T. F. 1904 The fate of the ovum and Graafian follicle in premenstrual life. Jour. Obst. and Gynec., vol. 5, pp. 1-12.

Stockarp, ©. R., AND PaPANICOLAOU, G. N, 1917 The existence of a typical oestrous cycle in the guinea-pig, with a study of its histological and physiological changes. Am. Jour. Anat., vol. 22, pp. 225-284.

STOECKEL, W. 1898 Ueber Theilungsvorginge in Primordialeiern bei einer Erwachsenen. Archiv. f. mikr., Anat., Bd. 53, S. 357-384.

STRASSMANN, P. 1923 Warum platzt der Follikel. Arch. f. Gynik., Bd. 119,

S. 168-206. THomson, A, 1919 The maturation of the human ovum. Jour. Anat., vol. 53, pp. 172-208.

1920 The ripe human Graafian follicle, together with some sug gestions as to its mode of rupture. Jour. Anat., vol. 54, pp. 1-40.

Van BENEDEN, E. 1875 La maturation de l’oeuf, la feeondation et les premiéres phases du développement embryonnaire des Mammiféres d’aprés les recherches faites sur la Lapin. Bull. Acad. Roy. des Sci., des Lettres et des Beaux-arts de Belgique, T. 40, p. 686.

VaN DER Strricut, O. 1908 La structure de l’oeuf de la Chienne et la génése du corps jaune. Compt. rend. Soc. d’Anat., 10° Réunion, Marseilles, p. 1.

1923 Etude comparée des ovules des Mammiféres aux differentes périodes de l’ovogénése d’aprés les travaux du Laboratoire d’Histologie et d’Embryologie de 1’Université de Gand. Arch. de Biol., T. 33, pp. 229-300.

Von Bakr, C. 1827 See Addison, ’27.

Watsy, L. 8. N. 1917 Growth of ovarian follicle—guinea-pig—under normal] and pathological conditions. Jour. Exp. Med., vol. 26, p. 245.

ZONDEK, B., AND S. ASCHHEIM 1927 Das Hormone des Hypophysenvorderlappens. Klin. Wochenschr., Bd. 6, S. 1321-1322. PLATE 1 EXPLANATION OF FIGURES

1 Ovum 5P, taken from a 4-mm. follicle from an ovary removed from a case of ectopic pregnancy of about the third month. A dense cumulus attached this ovum to the follicle wall. Many follicle cells were in mitotic division. The nucleus contained a large prominent nucleolus and evenly distributed chromatin material. The cytoplasm was uniformly granular. X 350.

2 Ovum 49P, removed from a 5-mm. follicle from an ovary which contained sixteen other follicles. The densely packed cumulus still contained mitoses. The nucleus was eccentric in position, but was otherwise normal. Several interesting cytoplasmie inclusions are shown in the section. x 350.

3 Ovum 42P, from a 7-mm., follicle from an ovary of a thirty-two-year-old woman removed on the ninth day of the menstrual cycle. The cumulus still attached the ovum to the follicle wall, but contained several lakes of secondary liquor folliculi. A fairly well-developed corona radiata surrounded the ovum. The nucleus was eccentrieally plaeed. The zona pellucida was stained darkly with iron alum hematoxylin. 350. (The photograph was made of the third of twelve sections.)

4 <A higher magnification of an adjacent section through the same ovum. It shows two mitotic figures in the two layers of follicle cells adjacent to the zona. This photograph also shows well the granular appearance of cytoplasm after fixation. xX 800,

5 <A normal ovum from a 10-mm. follicle removed on the twelfth day of the menstrual cycle. (The section taken does not pass through the nucleus.) This was chosen to show the texture of the cytoplasm and the close approximation of the zona pellucida to the vitellus without any intervening perivitelline space. Fight cells in mitosis were counted, indicating that the ovum was normal and the follicle still growing. X 350.

6 Ovum 35A, from a 9-mm. follicle removed on the twenty-third day of the menstrual cycle. This ovum was surrounded by a dense cumulus which still attached the ovum to the follicle wall. In twenty-five sections through this eumulus thirty cells were located which were undergoing mitotic division. Two are shown in the figure, one in the follicle cells adjacent to the zona at 2 o’clock, the other six to eight cell layers distant in the extreme Jower left corner of the photograph. Both were in the anaphase of mitosis. There is a slight indication of corona formation about this egg. > 350.

44 HUMAN OVA FROM LARGE FOLLICLES PLATE 1 EDGAR ALLEN ET AL.

45 PLATE 2 EXPLANATION OF FIGURES

7 A normal ovum from a 12-mm. follicle showing some evidence of coronaradiata formation. The nucleus was centrally located and contained a clearly defined nucleolus (in an adjacent section) and several chromatin masses. X 350.

8 Ovum 126P, from a 10-mm. follicle from the same ovary. This follicle was also growing rapidly, as shown by a high incidence of mitotic figures in the cumulus. One of these is shown in the upper center of the photograph. X 350.

9 Ovum 38A, obtained from a follicle 20 X 20 X 16 mm. in diameter removed on the twelfth day of the menstrual cycle. There had been no recent ovulation in this ease. The follicle was so large that on gross examinaticn it wes thought to be cystic. One of several lakes of secondary liquor folliculi is shown in the upper right-hand corner, an anaphase of cell division being located just to the left of this lake. There was a suggestion of corona formation. ~The nucleus was centrally located. The nucleolus had a clear center and a dark circumference. There can be little doubt but that this was a normal follicle still actively growing, but in spite of its large size the ovum had not yet begun maturation divisions. x 350.

10 The tenth of twelve sections through oocyte L17P, the only egg undergoing meiotic division which was recovered in this work. Twelve or thirteen chromosomes could be counted in the perivitelline space. These represent the first polar body. Portions of the cumulus are shown in the photograph. It is possible that some of the dispersion of the follicle cells may have been due to severe procedures in fixation and dehydration. X 350.

lland12 Sections 6 and 7, respectively, of the same oocyte. The plane of these sections cut through the chromosomes of the second maturation spindle which are seen at about 11 o’clock in these figures. A total of twenty-two chromosomes could be counted in these two groups. It was not possible to observe spindle fibers, due perhaps to a partial polar view of the chromosomes in an equatorial plate. X 350.

46 HUMAN OVA FROM LARGE FOLLICLES PLATE 2 EDGAR ALLEN ET AL. PLATE 3 EXPLANATION OF FIGURES

13 and14 Twin ova, 115P and 116P, from the same 10-mm. follicle removed from a thirty-six-year-old patient. The cumuli were composed of darkly staining pyenotic follicle cells. The zona pellucida was intact in both cases, but nothing was left of the cytoplasm but a suggestion of a coarse reticulum. X 350,

15 A degenerating ovum from a 6-mm. follicle removed on the eighth day of the menstrual cycie. The ovum was surrounded by a partly dispersed cumulus. The cytoplasm was extremely vacuolated and no nuclear structures were recognizable. X 350.

16 An atretic ovum from a 4-mm. follicle removed on the ninth day of the eyele. This ovum was floating free. The zona pellucida was frayed, the cytoplasm shrunken and vacuolated, and the nucleus in a very degenerate condition. A large globular nucleolus was clearly defined. 350.

17 An ovum in early atresia. A few pycnotic follicle cells in a partially dispersed state still surrounded the ovum. A large vesicular nucleus with prominent nucleolus was peripherally located. X 350.

18 An ovum from a 7-mm. follicle from the same ovary. Note the clear nucleoplasm, the ragged zona pellucida, and the dispersed cumulus containing many pyenotic cells. > 350.

48 HUMAN OVA FROM LARGE FOLLICLES PLATE 3 EDGAR ALLEN ET AL.

49 PLATE 4 EXPLANATION OF FIGURES

19 Ovum 17P, from a 6-mm. follicle. Compare figures 17, 18, and 19, which show three ova from follicles of about the same size from the same ovary. 350.

20 A degencrating ovum from a 6-mm. follicle from an ovary removed from a ease of ectopic pregnancy of the third month. Follicle cells of the cumulus were considerably dispersed and some were pyecnotic. The distribution of the material in the vitellus with an outer clear zone and a central coarsely granular mass is the special point illustrated. X 350. .

21 Ovum 37A from a 4-to-5-mm. follicle. This ovum was floating free, but was still surrounded by a considerable number of pyenotic follicle cells. Nuclear structures were clear-cut and appeared fairly normal. X 350.

22 An atretic ovum from a follicle 28 KX 18 X 16 mm. from an ovary removed on the eleventh day of the menstrual cycle. This ovum is somewhat distorted, a suggestion of a corona radiata is shown among the adjacent follicle cells. A portion of the nuclear membrane was still intact and a nucleolus was clearly visible. X 350,

23 A section of the wall of the follicle just described. There were six or seven layers of granulosa eells with a columnar layer bordering the theca interna. This arrangement of the outer layer of granulosa cells, as described by Robinson, is similar to corona formation about the ovum. X 350.

24 An ovum in early atresia. Note the dispersion of surrounding follicle cells, the ragged zona pellucida, and the irregular nuclear membrane—all clearly indicative of atresia. Xx 350. HUMAN OVA FROM LARGE FOLLICLES PLATE 4 EDGAR ALLEN ET AL. PLATE 5 EXPLANATION OF FIGURES

25 An ovum in extreme atresia obtained from a 4-mm. follicle. This ovum was very small, being cut into only three sections at a thickness of 104. The zona pellucida was distinct, but shrunken and distorted. There remained only a pale cloud of cytoplasm. The ovum had been invaded by several follicle cells. x 350.

26 Ovum 24P, from a 7-mm. follicle removed on the fifth day of the menstrual eyele. Interesting features illustrated were the invasion of follicle cells under the zona pellucida and the distribution of cytoplasmic elements of the vitellus in zones similar to that shown in figure 20. The nucleus was vesicular and the nucleolus appeared to be normal. X 350.

27 From a 5-mm. follicle removed during the fourth week after the onset of the previous menses. The zona pellucida was extremely swollen and tenuous and enclosed large numbers of cells, most of them probably follicle cells. The vitellus was still intact, no nucleus could be recognized, but several inclusions in the cytoplasm might possibly represent remains of chromatin material. X 350.

28 <A section through one pole of an atretic ovum from a 22-mm. follicle on the thirteenth day of the cycle. The interesting feature illustrated here is the crinkled nuclear membrane. A normal nucleolus was present in an adjacent section. xX 350.

29 An advanced stage of atresia in an ovum from a 4-mm. follicle on the fifth day of the menstrual cycle. Only a small part of the vitellus was recognizable. Many follicle cells had invaded the zona pellucida. X 350.

30 Extreme case of the same sort of atresia obtained from a 10-mm. follicle on the seventeenth day of the menstrual cycle. All that remains is a ‘ball of eells’ enclosed in a degenerate zona pellucida. This specimen is similar to oné type of ‘cell ball’ washed from the uterine tubes during our search for tubal ova. x 350. 53

PLATE 5