THE OESTROUS CYCLE IN THE RAT AND ITS ASSOCIATED PHENOMENA Memoirs of the University of California Vol. 6, pp. 1-148, 11 plates, 7 figures in text Issued June 28, 1922 THE OESTROUS CYCLE IN THE RAT AND ITS ASSOCIATED PHENOMENA BY JOSEPH A. LONG AND HERBERT McLEAN EVANS Memoirs of the University of California Editor, ARMIN O. LEUSCHNER l Volume 6 UNIVERSITY OF CALIFORNIA PRESS BERKELEY, CALIFORNIA 1922 CONTENTS PAGE I. Introduction 5 II. Literature 7 III. Methods , 12 IV. Gross Anatomy of the Reproductive Organs 14 V. The Normal Oestrous Cycle 17 A. Changes in the Vagina 17 1. Observations on the Living Animal 17 2. Correlated Histological Changes... 21 B. Changes in the Uterus 24 C. Changes in the Oviduct 26 D. Changes in the Ovary : 27 1. Ovulation 27 2. Relation between Oestrous and Ovulation 32 3. The Post-partum Ovulation 34 4. Corpora Lutea 34 Corpora lutea atretica sensu strictu 35 Corpora lutea with retained ova 35 Average number of corpora lutea produced at each ovulation 36 Size of corpora lutea of ovulation 37 Functional life span of corpora lutea ovulationis 37 E. Correlation of Changes in the Vagina, Uterus, Oviduct, and Ovary 41 F. Length of the Various Portions of the Oestrous Cycle 41 G. Total Length of Oestrous Cycle 45 H. Comparison of Oestrous Cycle in Rat and Guinea Pig 47 VI. The Attainment of Sexual Maturity 50 Age at First Copulation 50 Age at Establishment of Vaginal Orifice and at First Oestrous Changes in Vaginal Smear 50 VII. The Phenomena of Reproduction and Their Effect on the Normal Rhythm 54 A. Pregnancy 54 1. Length of Gestation 54 2. Proportion of Ova-producing Young 55 3. Effect of Pregnancy on the Reproductive Organs 56 Vaginal changes 56 Ovarian changes 58 Suspension of ovulat ion 58 Size and characteristics of the corpora lutea gravidates 63 Persistence of the corpora lutea gravidates 65 Effect of the corpora lutea gravidates on other corpora lutea 66 PAfJF B. Lactation 67 1. Effect of lactation upon the vagina 67 2. Effect of lactation upon the ovary 68 Suspension of ovulation 68 Corpora lutea lactationes 68 Effect of lactation on preceding corpora lutea gravidates 70 C. Copulation 70 1. Normal copulation and the formation of the vaginal plug 70 2. Infertile copulations 75 Diseased females with normal males r 75 Normal females with vasectomized males 76 3. Mechanical stimulation of cervical canal 76 4. Proof that copulation and cervical stimulation induce a condition of “ pseudo- pregnancy” 82 Effect upon the vagina 82 Effect upon the ovary 83 D. Production of Deciduomata 84 1. During normal ovulation cycle 85 2. During the pause following cervical stimulation 88 3. During lactation 89 4. Effect of ovariotomy 90 VIII. Effect of ablation and transplantation of various portions of the reproductive tract upon the normal oestrous cycle, pregnancy, and lactation 91 A. Hysterectomy 91 B. Excision of mammary glands 92 C. Ovariotomy 94 1). Ovarian transplantation 94 E. Exchange of ovaries between young and old animals 98 IX. Summary 100 Bibliography 104 Explanation of plates 115 Appendix 137 I. INTRODUCTION It lias long been known that sexual reproduction in the Mammalia takes place when the female is in a particular physiological condition known as oestrus. It has also been known since the time of Pouchet and Bischoff that in the ab- sence of copulation ovulation may nevertheless occur spontaneously at oestrus. In fact the only exceptions to this rule known at the present time would appear to be found in the cat and possibly in the rabbit. Oestrus and consequently ovulation are also known to be cyclic in many forms. There is no need to explain the profound importance of an infallible method for recognizing what is thus the periodic function of the ovary. The embryol- ogist. moreover, is interested in the precise time of ovulation inasmuch as knowl- edge of the sequence of events in the earliest development of mammals depends on such information. While in some animals oestrus is outspoken either by peculiarities of be- havior or by easily recognizable signs, in other forms this is not true. To the latter group belong some animals like the rat, of major scientific importance on account of their ready availability for experimental work. More than one method might be employed in attempts to recognize the oes- trous cycle. In lieu of various signs of oestrus, the chief acceptance of copulation—should serve as an indication of this time. In many mammals oestrus and ovulation occur on the day of parturition. In view of these facts, if access to the male were not permitted until the second day following partu- rition, the length of the oestrous cycle would be determined by the interval be- tween this post-partum oestrus and the next acceptance of coitus. Furthermore, on the assumption that the cycles are of uniform length, for example, ten days, coitus should take place not only on the tenth but also on the twentieth, thirtieth, etc., day after parturition. As a matter of fact, an attempt was made by one of us to determine the cycle in the rat and mouse on the basis of the above assump- tion. A large number of careful observations were made, but when these were analyzed no system pointing to a sufficiently uniform cycle could be discovered. Simplification of this method was then attempted in the use of vasectomized males by which it was sought to avoid the complications of pregnancy, but some of these complications were introduced by the mating act itself, as this paper will disclose. These studies gave as discordant results as did the former ones! In still another effort a series of animals was killed according to a schedule, and the ovary, itself, including the oviduct, thus examined in serial sections at precise and frequent time intervals after parturition. The present monograph will make evident how variations in the oestrous cycle itself and the inadequacies of method inevitably prevented recognition of the true cycle by the employment of any of the above devices. 6 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA For somewhat over thirty years scattered papers in the literature have indi- cated that periodic changes can be recognized in the structure of the mucosa of the generative tract in mammals, but no systematic study of these phenomena with a single recent exception seems to have been attempted. We refer to the papers of Moreau, La taste, Retterer, and Konigstein. Yet even though this older literature made it seem extremely probable that these changes were asso- ciated with the sexual history of the animal and in particular with oestrus, and were hence definitely related to ovulation, it was clearly impossible to follow these changes in any one animal, or even to predict when the}’' would recur, unless possibly advantage was taken of their relation to some outspoken event such as pregnancy. By the fortunate discovery that in the guinea pig these mucosal transformations are accompanied by the dehiscence of epithelial cells so that at times the lumen of the vagina has a characteristic cell content, it has been possible for Stockard and Papanicolaou to show us that we may discover with ease in the living animal the exact occurrence and progress of these cycles. When it has been proved, as Stockard and Papanicolaou have done for the guinea pig, and as we have been able to do with exactitude for the rat, that these cycles are correlated with the rhythmic discharge of ova from the ovary, it will be seen that Ave iioav have in our hands for the first time an accurate method for the detection of ovarian function in experimental animals. This fact promises important consequences, for it enables us to investigate disturbances of o Azarian function Avhich may be experimentally produced. Before proceeding to such studies, it Avould be necessary to establish clearly all the characteristics of AAThat AATe liaATe called the normal oestrus or reproductive cycle in the animal form which AA7e chosen to investigate. The present monograph is devoted to that study. We may remark at once that the rat, AAdiile exhibiting certain fundamental similarities AAdien compared Avitli the guinea pig, has also striking differences. It is hoped that this inquiry will serve as a pre- liminary to establish some of the more fundamental phenomena Avhich may be expected to characterize the various steps in the oestrous cycle throughout the higher mammalia. It has been carried on by us for a period of some four years and has been conducted with sufficient deliberation and repetition to lead us to cherish the hope that Ave have been able to establish reliable generalizations. It has immlved inquiries into the sexual behavior of about a thousand indi- vidual animals, upon each of which, for the period of observation or experiment, accurate daily records been kept. No one aat1io has not experienced a sim- ilar self-imposed, long-continued, and meticulous responsibility will readily appreciate the amount of conscientious concern necessary to such tasks. Well over five hundred careful autopsy protocols liaAre been founded on the study of complete serial sections of ovarian and representative sections of Anginal material. Grateful acknowledgment is made of the grants awarded by the Board of Research of the University of California and the American Medical Association for the prosecution of this work. LONG-EVANS: OESTROUS CYCLE IN THE RAT 7 II. LITERATURE In view of the excellent work of Marshall on the Physiology of Reproduction, a comprehensive review of the very extensive literature on this subject is un- necessary. However, for convenience, a list is presented of the times of oestrus and ovulation which have been reported for various mammals by various observers: Animal Length of Dicestrous cycle Ovulation Authority Carnivora : Cat Dependent on copulation Coste (Hansen) 1847 14 days Marshall & Jolly 1905 Only at copulation f Ancel & Bouin 1909 [Bouin & Ancel 1909 Fraction of gestation Only after copulation and inde- pendent of abortion Winiwarter & Saimont 1909 About 14 days Marshall 1910 Only after copulation Longley 1910-11 Only after copulation Van der Stricht, R. 1911 Only after copulation Hammond & Marshall 1914 Dog Spontaneous Spallanzani (Heape) 1786 Spontaneous at heat Bischoff 1844-5 Spontaneous Sir E. Millais 1884 Spontaneous Pierre Rossi 1884 Spontaneous (?) at heat Iwanoff 1900 Spontaneous at oestrus Spontaneous and associated with Marshall & Jolly 1905 rut Ancel & Bouin 1908 Spontaneous Ancel & Bouin 1909 Spontaneous Marshall 1910 Spontaneous at oestrus Hammond & Marshall 1914 Lioness 3 weeks Marshall & Jolly 1905 Otter Month Marshall & Jolly 1905 Marshall 1910 Marsupialia : Dasyurus Independent of copulation Hill 1910 Spontaneous, after oestrus Hill & O’Donoghue 1913 Didelphys virginiana Spontaneous Hartman r 1916 (1919 Rodentia: Ferret Usually only with copulation, al- though sometimes spontan- eous at oestrus Marshall 1904 8 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Animal Length of Dioestrous cycle Ovulation Authority Guinea Pig Irregular Spontaneous within 24 hrs. pp. Bischoff 1852 38, 43 & 44 days During first 24 hrs. pp. May be (Hensen) spontaneous, but greatly in- fluenced by copulation Reichert 1861 Spontaneous, but may also b; influenced by copulation Bischoff 1870 18 (17, 18, 35, 37) Spontaneous, also 6-10 hr. pp. . Hensen 1875 days Ovulation and copulation co- incident Rein 1883 15-20 days (?) Retterer 1892 10-20 days At rut Lataste 1892 Spontaneous Iwanoff 1900 Spontaneous after parturition Lams & Doorme 1905 5-20 hrs. pp. Vagina open only at heat, and ovulation when vagina opens Rubaschkin 1905 Spontaneous after parturition Sobotta 1906 18 j-24 days Spontaneous 6-10 hrs. pp. Loeb, L. 1909-11 Only at copulation Bouin & Ancel 1909 Only at copulation Ancel & Bouin 1909 Not independent, exceptional without copulation Bouin & Ancel 1910 Spontaneous (?) 12-17 hrs. after parturition and 2-4 hrs. after copulation Lams 1913 16 days Loeb & Hesselberg 1917 16 days Spontaneous Stockard & Papanicolaou 1917 Mouse 10 (or 20) days Spontaneous. Copulation only after ovulation Lataste 1883 Copulation at fairly equal intervals Tafani 1889 10 days Morau 1889 10 days At rut. Spontaneous (?) Lataste 1892 21 days Spontaneous Sobotta 1895 10 days Heape 1900 Spontaneous Lams & Doorme 1905 Spontaneous Gerlach 1906 Oftener than once Spontaneous Melissinos 1907 during (21)30 days Spontaneous after parturition Kirkham 1907 Only at copulation Bouin & Ancel 1909 Spontaneous after parturition Kirkham 1910 Spontaneous after parturition Long & Mark 1911 days Spontaneous Long & Smith 1916 163-19 days Spontaneous Smith 1916 LONG-EVANS: OESTROUS CYCLE IN THE RAT 9 Animal Length of Dicestrous cycle Ovulation Authority Rabbit 72 hrs. after copulation De Graaf Eggs in oviduct 3 da. p. coitum Cruikshank 1797 9-10 hrs. after copulation Barry 1839 38, 43, 44 days Spontaneous or 9-10 hrs. p. coitum 9-10 hrs. after copulation Spontaneous after parturition May be induced by copulation Bischoff Reichert \ Weil 1842 1861 1873 35, 37 days Induced by copulation About 12 hrs. after copulation > Hensen 1875 8-10 hrs. after copulation Spontaneous at heat immediately Van Beneden 1880 after parturition Ott 1882 No periodicity Ovulation and copulation coin- cident Rein 1883 15-30 days (?) Retterer 1892 About a month? Lataste 1892 Only at copulation, 10 hrs. after Heape '1897 . 1905 Spontaneous pp. Iwanoff 1900 10-21 days. Average 15 days. Very vari- able Heape 1900 7-10 hrs. after copulation Regaud & Dubreuil 1908a 30 days ? Occasionally spontaneous. Pre- sence of male, without copula- tion does not cause ovulation Regaud & Dubreuil 1908c Only at copulation Bouin & Ancel 1909 Only at copulation Ancel & Bouin 1909 Shortened by presence Only at copulation. Hastened of male by mating Dubreuil & Regaud 1909 10-15 days (Heape) Usually only at copulation, but even three weeks sometimes spontaneous Hammond & Marshall 1914 Rat, white Spontaneous at rut? Bischoff 1844 10 days Morau 1889 10 days At rut, spontaneous Lataste 1892 1 ,1893 10 days Heape 1900 14 days Konigstein 1908 Spontaneous after parturition Burchard 1910 Spontaneous after parturition Kirkham 1910 brown 5 days Miller 1911 white Spontaneous after parturition King 1913 21 days (?) Spontaneous Kirkham & Burr 1913 10 days Spontaneous Long & Quisno 1916 5 days (average) Spontaneous Long & Evans, present paper 1922 Dipodillus 8, 9-10 or 14 days At rut Lataste 1883 Rodents in Only at copulation Ancel & Bouin 1909 general 10 days, even 5 days At rut Lataste f 1887 (1893 Muridae 110 days At rut Lataste 1892 (rats, mice, Gerbillus, Meriones) Muridae 10 days Lataste 1887 (Eliomys, Gerbillus, Dipodillus (2 sp.), Meriones (2 sp.) 10 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Animal Length of Dicestrous cycle Ovulation Authority “Laboratory rodents ’ ’ 3-4 weeks Konigstein 1907 Ungulata: Cattle 3-4 weeks Ellenberg 1892 19 days in summer 20-21 days in winter Wallace f 1876 (1904 Cow 3-4 weeks Fleming 1878 Spontaneous, at heat Iwanoff 1906 Spontaneous Bouin & Ancel 1909 21 days Williams 1909 Spontaneous at oestrus Marshall 1910 Spontaneous at oestrus Hammond & Marshall 1914 2nd day of 21 day cycle Spontaneous Kupfer 1920 21 days Spontaneous Zietzchmann 1921 Sheep Only after copulation Hausmann 1840 Spontaneous at rut Bischoff 1844 a, b 2-4 weeks Fleming 1878 17 days or 3-4 weeks With or without (?) copulation Bonnet 1884 20-30 days 15-16 days Spontaneous at earlier periods Ellenberger 1892 but not at later. Hastened by copulation Marshall 1901-3 Spontaneous at rut Iwanoff 1906 Sheep Lombardy 13-21 days Marshall 1910 Spontaneous at rut Hammond & Marshall 1914 Sow After copulation Hausmann (Marshall) 1840 Spontaneous at rut Bischoff 1844 2-4 weeks Fleming 1878 Spontaneous Bouin & Ancel 1909 2-4 weeks Spontaneous, at oestrus (probably) Marshall 1910 Spontaneous at rut (certain) Corner & Amsbaugh 1917 18-23 days (mean 21 4-9 days pp. Struve 1911 days Spontaneous at rut Corner 1917 21 days Spontaneous Corner 1921 Wild cattle, deer, and other similar forms in captivity 3 weeks Heape 1900 Donkey Spontaneous at heat Marshall 1910 Mare 2-4 weeks Fleming 1878 Spontaneous Heape 1897 Spontaneous at heat Iwanoff 1906 Hastened by copulation Dubreuil & Regaud 1909 Spontaneous Bouin & Ancel 1909 About 3 weeks Spontaneous at oestrus Marshall 1910 Spontaneous at oestrus Hammond & Marshall 1914 All animals Spontaneous Pouchet 1847 LONG-EVANS: OESTROUS CYCLE IN THE RAT 11 It will be noticed that ovulation is spontaneous at rut in all except the cat, ferret, and rabbit, and, in the latter two, spontaneous ovulations are recorded. The list shows also that, while the oestrous cycle is known in some of the larger animals, and in some cases the exact relation between oestrus and ovulation, it is in the guinea pig only, among laboratory animals, that the oestrous and ovu- lation cycle has been carefully worked out. It is interesting to note the ten-day cycles given for the rat and other rodents by several investigators and especially the variable period given for Dipodillus by Lataste. “The Changes that Occur in the Non-Pregnant Uterus during the Oestrous Cycle” forms the subject of a chapter of Marshall’s Physiology of Reproduction, which summarizes the researches on the larger mammals including man. Very little space, however, is given to the rodents, not even his own work on the ferret being cited. Probably lack of room prevented any mention, more than a single sentence, of vaginal changes, especially since most published accounts deal with the pregnant animal. There are, however, several papers on changes in the vaginal mucosa at rut and during pregnancy. In the earliest of these, Morau (1889) describes the vaginal mucosae, occur- ring at rut, of mice killed at intervals after copulation. The periods of rut recur at intervals of ten days except during lactation and pregnancy. At the time of copulation or rut the epithelium is stratified and consists of three distinct parts, the outermost of which is cornified. By the fourth day after copulation the cornified layer has become loosened and completely lost. If the animal is pregnant, the cornified layer is apparently not re-formed until the next rut after parturition. In animals which fail to become pregnant, the condition at copulation may recur on the eighth and sixteenth days, and one case is recorded of desquamation on the fourteenth day. Morau states that he finds the same conditions in the rat (both black and albino) at rut, and that the vaginal epithe- lium undergoes changes in the guinea pig, rabbit, one of the species of Gerbillus, and all rodents. Although Morau’s work was done primarily on pregnant mice, it is clear that he has observed a cyclic change which is independent of preg- nancy. His observations being confined to sections, it was not possible for him to follow the cycles in one animal. Like Lataste, he thought that the desqua- mated epithelium formed a “vaginal envelope.” In 1892 and 1893 several papers were published by Lataste and by Retterer on the periodic transformations of the vaginal epithelium of rodents. Ret- terer ’s first paper is concerned chiefly with the conditions of the epithelia in the dog, cat, sheep, and pig during pregnancy. He thinks that rut and copu- lation do not influence the mucosa, but that the last days of gestation and partu- rition do in carnivora and ruminants. In a second paper (1892b) Retterer describes the changes occurring in the vaginal mucosa of the non-pregnant guinea pig. The animals were killed at 12 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA intervals during the first twenty days following parturition. For the first fif- teen days the superficial cells are mucous, hut at the fifteenth day there is formed under the superficial layers a cornified stratum which is easily detached during manipulation. A similar cornification occurs in the dog at rut. In the cat it is stratified but not cornified, while in the non-pregnant rabbit the mucosa becomes greatly thickened and stratified. Lataste (1892-1893) points out that his own and Morau’s work show that the vaginal changes are associated with rut primarily rather than with gesta- tion. In rodents (rats, mice, Gferbillus, Meriones) at the approach of rut the lips of the vagina become swollen so that it is possible to tell a day before when a female will copulate. This swelling also corresponds to a cornification of the mucosa. After rut the keratinized layers are lost, forming a vaginal envelope or, if a plug is present, the outer parts of the plug, as described in Lataste’s papers. Lataste suggests that the cornification is a protection especially at the occurrence of copulation. There is, then, a rhythm in the vaginal mucosa, a temporary thickening at rut followed by a thin mucous-like epithelium during the interval. Konigstein (1907) describes the changes in the genital tract during preg- nancy and at heat in several rodents used in the laboratory (rats, guinea pigs, and rabbits), also in a dog. Although he deals with changes in the vagina during pregnancy chiefly, he clearly indicates that in the non-pregnant rat the flattened epithelium is transitory, and that it is completely cast off and a new one formed four days after parturition. Such flattened epithelium may contain leucocytes even to the outer layers. Apparently it is gone after heat and is replaced by cylindrical epithelium. The most recent papers on the subject of oestrus are the excellent ones on the guinea pig by Stockard and Papanicolaou (1917, 1919). They were able, by inspection of the vagina and by microscopic examination of smears, to deter- mine the time of oestrus and ovulation, and the length of the cycle. They de- scribe also the histological changes in the vagina, uterus, and ovaries. These are compared in some detail farther on (p. 48) with those in the rat. III. METHODS A satisfactory conception of physiological phenomena in this field cannot be obtained from a miscellaneous material of unknown or greatly varying age and of uncertain past sexual history. In such studies the greatest emphasis must be laid on the necessity of securing individuals in full reproductive vigor, conditions which are obtained with most certainty, of course, by the establish- ment and careful control of an animal colonv. LONG-EVANS: OESTROUS CYCLE IN THE RAT 13 The rats used in this study were both white and colored, descendants of a cross made about seven years ago between several white females and a wild gray male caught in Berkeley, black, gray, and hooded varieties resulting. It is a vigorous stock due in part, no doubt, to the cross, for one of the original hybrid females had her last litter when nearly two and one-half years old. As far as our experience goes, there is no difference to be observed between these different colored rats with respect to the oestrous cycle. The colony has been fed with table scraps supplied daily from a large hotel, thus insuring them a variety of nutritious food. In the case of animals used for experimental purposes this diet was supplemented by whole milk and occa- sionally by other substances, such as uncooked liver, greens, etc. Although the recent wealth of studies in nutrition makes it no longer necessary to speak of the necessity of a generous and varied diet of both the correct nutritive and vitamine consistency, we may be pardoned for emphasizing it by at least this brief mention because we have been able to satisfy ourselves of the effect of a poorer nutritional regime in consequent irregularity of the sexual cycles. It may be questioned whether table scraps of notorious variability in food value could prove an ideal food. One should eliminate the fluctuations which will characterize any but well standardized diets. We feel that it is necessary also to dilate upon the great importance of daily and regular care, rigorous cleanliness, and access of the individual animals to adequate space and air. We have observed repeatedly in attempts to rear con- siderable numbers of animals in single cages that, though the space seemed more than adequate and the ventilation through open meshwork all that could be desired, the gregarious habits of the rodent actually prevented these needed desiderata from being obtained. These animals tend habitually to sleep in piles, even in warm rooms where the need for conserving body temperature could not exist. Though it may seem doubtful that we could explain the poorer state of a colony maintained in this way by respiratory inadequacy, it is never- theless true that separation of animals into pairs or at most into fours in smaller unit cages has in our hands proved to be the only dependable method of insuring vigorous stock. It may be thought also that these rodents would be able to maintain perfect health in the presence of very considerable fluctuations of temperature, represented by diurnal or seasonal climate, but in our opinion such is not the case. They thrive so much better under conditions of equable tem- perature that we have taken the pains to install an electric source of heat and, providing the appropriate supply of fresh air is at hand, we have seen only good effects with the maintenance of a temperature from 65° to 68° F. This is especially true where animals are “naked” in the cages, so to speak, that is, where one does not install elaborate nest materials and separate nest boxes— things of great nuisance to handle and to* clean and which we have learned to dispense with entirely. 14 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA A first hand acquaintance with rats of some years’ duration has emphasized to us further the inestimable advantages of direct and frequent handling of his stock on the part of the investigator. We do not hesitate to refer the formid- able attitude of aggression on the part of this animal, of which bacteriologists and others speak, to its manipulation with tongs or other rough devices. If grasped with the hand so that the head may not he turned to inflict a bite and if handled gently, these animals may be controlled with surprising ease. We do not indeed handle them in any securer fashion in inflicting needle wounds in intraperitoneal injections of vital dye, and we dwell upon the matter because the infliction of pain or of unnecessary fright and the consequent retaliation re- sulting therefrom might lead other investigators to give up the hope of pur- suing that intimate acquaintance with the individual animal which is demanded in an investigation of this sort. An unaided observer will have no difficulty in holding a gentle animal with one hand and with the other introducing a small speculum into the vagina or taking samples of the vaginal content by means of the introduction of a pipette, spatula, or other instrument. No natural com- bativeness on the part of the rat is known to us except the defense exhibited by a suckling mother. The female with young will not hesitate to protect her brood, but, if driven from them—and her tendency to alarm makes this easy—she may be grasped by the tail and removed from her cage. Once out of the accustomed surroundings and mastered by being grasped, as is usual, behind the shoulders, a speedy change in her psychology ensues and she may be handled further with perfect impunity. A final word on method may be allowed in the description of a device (fig. A) for preventing suckling of a newborn brood. We have found it expedient to place the mother in an “obstetrical” compartment with an inclined or slanting floor which has a small opening at its lower edge. In our experience, this device does not permit much time to elapse before the newborn young roll out through the slit arranged at the lower end. Normally, even if the period of suckling has been only a few hours long, this nevertheless shows itself in the young in the peculiar white spot created by the milk-filled stomach which shimmers faintly through the body wall. Young delivered from the floors of these compartments do not appear to have succeeded in suckling at all. This device was essential for us, as will be shown later, when we sought to establish the effect of lactation on ovulation and on the corpora lutea. IV. GROSS ANATOMY OF THE REPRODUCTIVE ORGANS The essential relations of the reproductive organs in the rat are shown dia- grammatically in figure 1. The ovary lies within a small, completely enclosed space, the periovarial space, the walls of which are formed chiefly of the perio- varial membrane. The periovarial space has no connection with the peritoneal LONG-EVANS: OESTROUS CYCLE IN THE RAT 15 Sliding top. Spring of curved sheet brass. Section through cage with sloping floor of J inch wire cloth. Top, back and sides of sheet metal. ..Metal strips. Weight.' .Torsion wire to start swinging arm. .Opening through which young roll to fall into box. Top view with box removed. Tray for catching refuse. adjustable counter weight of sheet lead. Box. Box. Pirn , Pivot Weight. .Spring of curved sheet brass... ..Swinging arm of wire \Torsion wire to start swinging arm. Side view. Front view. Fig. A. Apparatus to prevent newborn rats from being suckled and to record automatically the time of their birth. This figure represents the essentials of one unit of the apparatus, part of which is a chronograph with a drum revolving once in 24 hours. As many units may be operated as desired. The pregnant female is placed in the cage, which offers ample room and from which she is removed daily for feeding. Since nest material cannot be used, the cage must be kept in a room sufficiently warm. The young as they are born roll out and drop into the box before the mother can suckle them, for without constant attention the young rats cannot be kept in the cage. The box rests on a balance so adjusted that a small addition of weight to the box depresses it in the direction of the arrow. The movement thus imparted to the balance raises the pin and releases the swinging arm which in its downward movement scrapes along the edge of the spring of curved sheet brass, making an electric contact. The current thus closed for a fraction of a second operates an electromagnet which through the medium of a pointer records its movement on the smoked paper of the drum. It is then a simple matter to determine the time of birth of the young rats (or mice). In order that the swinging arm shall start without fail it is attached to one end of a fine spring brass wire, the other end of which terminates in a narrow rectangular loop. The loop passes over a post and is adjusted at such an angle to the arm that when the arm is raised into position the torsion wire is twisted enough to keep the upper end of the arm against the pin and on release to propel it through a part of its arc of motion. The torsion wire also serves to keep the arm in contact with the spring of sheet brass. X ca. 16 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA cavity as shown by the distention of the membrane by a characteristic fluid at the time of ovulation. The oviduct is thrown into eight to ten folds which may be divided into two groups, one the distal, the other the proximal group. The proximal group, according to Huber, is capable of still further subdivision as Ovary Oviduct Right horn of uterus Cervical canal - Cervix Vaginal portion of cervix Vagina „ External orifice of vagina Fig. 1. Semidiagrammatic ventral view of the reproductive organs of the female rat partly laid open by a cut in the frontal plane. The lumina of the two uterine horns communicate with the vagina independently through the two cervical canals which open in the tip of the cervix protruding into the vagina. For details of oviduct and ovary see figures 38 and 39, pages 27 and 28. Sobotta lias described for the mouse. The distal group communicates with the periovarial space through the club-shaped, ciliated end. The folds of the distal group differ from those of the proximal in being very freely movable, in having LONG-EVANS: OESTROUS CYCLE IN THE RAT 17 thin walls allowing of great distention, and in having the mucosa thrown into high ridges and covered with distinctly ciliated epithelium. The proximal part of the oviduct does not open into the exact tip of the uterine horn, but, as Fiscliel has described, through a papilla slightly lateral to the tip. Although the walls of the uterine cornua are fused in their lowermost por- tions, their lumina are entirely independent of each other, opening into the vagina through separate orifices. These external orifices of the cervical canals are difficult to detect among the lappets of the vaginal portion of the cervix. Their position is indicated by bristles in figure 2, plate I. V. THE NORMAL OESTROUS CYCLE A. Changes in the Vagina 1. OBSEBVATIONS OX THE LIVING ANIMAL If the vaginal mucous membrane of rats in full reproductive vigor is ex- amined daily by means of an appropriately small speculum, two strikingly different conditions will be encountered from time to time. These conditions, which alternate with each other, may be roughly described as the “ moist, pink- ish,” and as the “dry, white” conditions, and either condition will usually be found to recur in four days. With the short cycle of four days, which we have found to be normal for the majority of young adults, it accordingly happens that during about one-lialf of this time the vagina has the moist, pinkish ap- pearance characteristic of the dioestrous pause and during the other half the “dry” condition associated with a succession of events grouped together and related to oestrus proper. The moist, pinkish condition typifies the dioestrous pause, or the interval between oestrous changes. The dry, lusterless, or white condition is usually associated with a swelling or turgescence of the small radi- ating folds about the vaginal aperture (fig. 3, pi. I). In its incipiency it is always associated with the manifestation of oestrus, and toward its close witli the occur- rence of ovulation. Vaginal Smears. These stages succeed one another in orderly sequence, and each is charac- terized by a different histological make-up of the vaginal fluid. It is indeed not the least remarkable of the histological phenomena encountered in these studies that we should have so clearly marked a succession of the cell types which are thrown off within the vagina. Our study has given us a satisfactory explanation of the origin of these cells and has also in a sense made it clear why we should have this succession, but they have only served to make more precise our acquaintance with the astonishingly orderly and steplike character of the 18 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA cellular dehiscence, by means of which there appears in the vaginal lumen at any one time only one type of epithelial cell. These studies have shown us that the dehiscence of the epithelium, which proceeds relatively speedily, also takes place at about the same rate over the entire mucosal surface, for were it other- wise we would not have the homogeneous cell picture which we actually find, but only confusion arising from the occurrence in some localities of an earlier, and in others of a later, succession of the same events. We have chosen the stages disclosed by change in the content of the vaginal smear as of fundamental value in our study—as the stages with which to cor- relate changes in the histological structure of the internal organs of reproduction for two weighty reasons: (1) In contrast to other phenomena (e.g., the appearance of external swell- ing and, in particular, the duration of oestrus), each of the stages marked by changes in the vaginal smear is approximately constant in length in all animals. (2) The “vaginal stages” constitute the only reliable method of recognizing subdivisions of the oestrous cycle in the living animal. We have, accordingly, made the histological characteristics of the vaginal smear the criteria for delimiting steps in the oestrous cycle and for this reason, before giving an account of the changes undergone by the generative apparatus in each stage, shall begin with a description of the successive histological pic- tures found at each stage of oestrus in the vaginal smear. (See table 1, p. 42.) Dioestrous Interval. During the interval, or dioestrous pause, which constitutes about half of the entire cycle and which possesses a relatively clean, moist, glistening, normally somewhat translucent, pinkish mucosa, the pipette, spatula, or other sampling instrument will succeed in withdrawing from the vaginal lumen a variable quan- tity of thin, somewhat stringy mucus in which are entangled leucocytes and small, irregularly shaped, free epithelial cells (figs. 4 and 5, pi. II). Both types of cell may occur in considerable numbers in some animals or may be both of them very sparse in others. The leucocytes are usually fairly abundant and are the characteristic small, polymorphonuclear elements which, as is well known, often have annular nuclei in the rat. The leucocytes are often distorted in one dimension so as to be slightly elongated within the “strings” of mucus. The epithelial cells are always single, never in groups. Stage One.—Oestrus is inaugurated by the occurrence of a distinct stage which we have designated as “Stage One” (the stage corresponding to Heape’s pro-oestrum). While characterized by a distinctive histological picture of the vaginal content, it may also often be detected macroscopieally by the changed appearance of the surface of the vaginal mucosa as disclosed by the speculum; LONG-EVANS: OESTROUS CYCLE IN THE RAT 19 the mucosa is no longer moist, seems distinctly less transparent, and, while pos- sessing a slight sheen, or luster, has a characteristic opaque look. In some ani- mals a further external characteristic of Stage One may be found in an increas- ing turgescence of the small radiating folds about the vaginal aperture (fig. 3b, pi. I) as compared with the condition during the dioestrous interval (fig. 3a). To the naked eye. hoAvever, these appearances are not so distinctive that they may not be overlooked, but under no conditions will the incidence of Stage One escape the observer who takes the precaution to obtain a microscopic sample of the contents of the vaginal lumen. To a small drop of physiological saline, Ringer’s, or Locke’s solution, a narrow spatula may be applied after its withdrawal from the vagina where it has been in contact with the mucosal surfaces. This simple method of sampling the vaginal content is actually adequate for an infallible diagnosis even when studied with a lens Avhieh gives a magnification of but eighty diameters. The small translucent, jelly-like, adhesive mass withdrawn on the tip of the spatula is not easily dislodged into the drop of saline solution upon the slide. In it leucocytes have entirely disappeared and great numbers of small, round, nucleated epithelial cells of strikingly uniform appearance and size are noAv present (figs. 10 and 11, pi. II). Occasionally small sheets of these cells may be encountered. The microscopic picture is absolutely characteristic and unmistakable, for these particular cellular elements have occurred and will occur at no other time in the oestrous cycle. The manner of their formation as a peculiar layer at the surface of the vaginal epithelium will be disclosed later. Not the less striking and unexplained is their sudden dehiscence, a fact corre- lated with the equally sudden cessation of leucocytic migration. Leucocytes, indeed, are not encountered again until the close of the next two stages in oestrus. Stockard has found a similar behavior on the part of these cells during the pro- oestrum and oestrus in the guinea pig and it Avould appear that this peculiar lull, or sharp pause, in the penetration of the mucous membranes by leucocytes, which is otherAvise going on constantly throughout the dioestrous pause, is a very fundamental characteristic of the histology of oestrus in the tAvo rodents which have been investigated. The average duration of Stage One is twelve hours. During Stage One females will usually not accept copulation, defending themselves against any aggression on the part of the males. It is hence proper to recognize this as the stage which is normally preparatory for oestrus. In some cases, however, at about the middle of the stage, and in still more cases toward the end of it, animals will mate, but even under these circumstances they do not usually show oestrous excitement in the characteristic manner typical of the folloAving stage. As aauII be described at some length beloAv, the accept- ance of coitus can easily be tested by employing a small number of young accus- tomed males in a well illuminated, flat cage which can be instantly opened in order to interrupt the act. 20 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Stage Tivo.—In Stage Two the macroscopic changes which we had noted as characterizing Stage One, notably the beginning swelling of the vaginal lips and the dry mucosa, are now increasingly evident. The speculum or spatula meets notably more resistance at attempted introduction into the vagina. An equally definite change takes place in the microscopic character of the vaginal smear, for the small, nucleated, and somewhat granular appearing epithelial cells char- acteristic of Stage One are rather suddenly replaced by large, thin, transparent, non-nucleated, scalelike elements—the cornified cells (figs. 21 and 22, pi. II). For this reason Stage Two may well be designated as the “cornified cell stage.” The sample, while still scanty in amount, is opaque, whitish, and granular, the crumbling particles being easily tapped off* into the drop of saline, where they have a tendency to float. There is still a singular absence of leucocytes, whose sudden disappearance was so marked a characteristic of Stage One. It is during the beginning of this stage that the female usually shows unmistakable signs of heat. If placed within a mating cage and, through some chance, not approached by the males, she will manifest what we have been led to recognize as oestrous excitement by quick, darting motions, or hops, with the back arched, with occa- sional quivering of the body and with a curious shaking of the ears. We may state that this behavior, never encountered at any time other than oestrus, need not. however, invariably be manifested by individuals in heat. A simple and unmistakable test of heat is 'furnished by the attitude of the female on the ap- proach of the males. Mating is made possible by a characteristic flattening of the back, or slight opisthotonos. During this period there is usually a dis- agreeable odor attached to the vaginal secretion, which perhaps is a means of attracting and exciting the male. Stage Three.—Stage Three, which may be equally well named “the late cor- nified cell stage,” cannot be separated abruptly from Stage Two, the histological picture of the vaginal smear being identical although an exaggeration of that characterizing the preceding stage. Indeed, the accumulation of cornified, non- nucleated epithelial plates within the lumen of the vagina now proceeds so rap- idly that easily visible masses of whitish, granular, or pasty substance always occur deep in the vagina near the cervix. These masses consist exclusively of enormous numbers of the elements in question without admixture with other cells. One avIio is familiar with the character of the “plug” (bouchon vaginale) which the male rat leaves in the vagina after coitus will at first easily mistake these white masses for fragments of the “plug”. The most superficial micro- scopic examination, however, shows that they are made up of sheets or of single cornified elements whereas the male secretory product is amorphous. But be- sides this accumulation of cheesy substance, macroscopically evident within the vagina, Stage Three is further typified by a very important characteristic in the LONG-EVANS: OESTROUS CYCLE IN THE RAT 21 fact that animals in this stage will usually no longer accept coitus.1 The average combined length of Stages Two and Three would appear to he about thirty hours. Throughout this period the vagina is dry.2 The swelling of the vaginal orifice may also persist. Stage Four.—Stage Four, the metoestrum or the leucocyte-cornified cell stage, is inaugurated by the appearance in the vaginal smear of leucocytes among the cornified cells (figs. 31 and 32, pi. II) and ends with the disappearance of the latter. The leucocytes cause a softening of the granular masses seen in Stage Three and convert them into a substance of a cheesy, creamy, and increas- ingly fluid consistency. Before the cornified cells completely vanish from the smear, epithelial cells reappear so that during a short interval all three cellular types are present (cornified, non-cornified, and leucocytes). This ushers in the beginning of the dioestrous pause, which may be recognized by the complete disappearance of cornified elements so that the smear again consists of leuco- cytes and epithelial cells. Stage Four is normally of about six hours duration and might well be known as the stage of transition to the resting condition. It is chiefly characterized by the sudden resumption of leucocytic migration. It is interesting that in a certain proportion of animals the epithelial des- quamation of the mucosa may proceed somewhat further than to the cornified cell layer before leucocytes come in, and hence we may have the cheesy state, Stage Three, immediately succeeded by a few hours of a stage in which rather large, spindle-like, nucleated epithelial cells are shed. In such cases, however, vestiges of the cornified cell are always present and when leucocytes appear three cell types are actually found. Since there is thus an invariable succession of cell types occurring in the vaginal smear at various times in the oestrous cycle, the stages in the cycle may with equal propriety be named from the cell content of the vaginal smear. Should we do this, we could designate them as the stage of the sudden appear- ance of masses of uniform sized, nucleated epithelial cells dehisced from, the sur- face, the stage of few large cornified cells, the stage of extremely abundant cor- nified cells, the stage of many leucocytes admixed with cornified cells, and, finally, in the dioestrous pause, the stage of leucocytes with scanty epithelial cells. 2. CORRELATED HISTOLOGICAL CHANGES Histological changes in every portion of the reproductive tract are corre- lated with the above stages which we have characterized by means of the vaginal smear (see table 1, p. 42). For obvious reasons we shall begin with a descrip- tion of the histology of the vaginal mucosa, the changes in which explain so 1 However, see the description of length of oestrus on page 33. 2 Except for the very brief period of time which may immediately succeed sudden emptying of the uterus and drainage of its fluid into the vagina. 22 MEMOIRS OP THE UNIVERSITY OP CALIFORNIA beautifully the characteristics of the vaginal smears. Our description refers to the histological conditions occurring in the mucosa of the vaginal folds at, or near, the cervix for the reason that: (1) There can be no uncertainty concerning the position of the region in question. (2) Possible injury caused by examination with the spatula is least likely to occur so deeply. (3) It must be confessed that, in spite of the remarks which have been made previously regarding the singular, synchronism in epithelial dehiscence, the changes do not actually take place at exactly the same rate or to the same degree in all parts of the vagina, even though they tend to do so. To begin with, we may summarize the main histological events occurring in the vaginal mucosa by stating that the mucosa is depleted to a very low layer during the resting period, or dioestrous pause, the epithelium grows rapidly as Stage One (the pro-oestrum) approaches, and with its growth undergoes a dif- ferentiation into (1) a surface layer, (2) a subjacent stratum corneum, and (3) immediately subjacent to the latter, a small, but distinct, stratum granu- losum beneath which are (4) several cell layers of the rete mucosum, and finally (5) a basal layer or stratum germinativum. This highly differentiated struc- ture falls asunder during the active stages of the oestrous cycle (stages 1, 2, 3, and 4) by successive dehiscence of its various layers. Dioestrum. During the dioestrous interval the vaginal mucosa (fig. 6, pi. Ill) is thin, consisting of from four to seven simple cell layers, the squamous transformation of its upper cells being only slight. It remains constantly infiltrated by a certain number of leucocytes and constantly decreases in thickness through dehiscence of its superficial cells, a process evidently not completely compensated for by the scanty mitoses in the basal layer. Near the end of the dioestrous pause, however, these mitoses increase greatly in number and dehiscence almost ceases (fig. 7, pi. Ill). The epithelium, consequently, becomes higher (8 to 9 cells) and a squamous transformation of it becomes expressed. The actual surface cells, however (constituting a layer one to three cells deep), do not become cornified but become transformed in a characteristic way by swelling and other changes so as to form a well defined superficial layer, the layer of first stage cells. Co- incident with this, squamous changes have advanced and immediately beneath the surface layer noticeably large, flattened, though nucleated elements are found. These are destined to undergo a further transformation into cornified cells, as later stages disclose. LONG-EVANS: OESTROUS CYCLE IN THE RAT 23 Stage One.—In Stage One, or the pro-oestrum, the peculiar surface cell layer which we have described and which is unaffected by squamous transfor- mation now reaches its fullest expression (figs. 12, 13, 14, pi. Ill), and begins to separate from the subjacent layers which themselves have undergone a strik- ing differentiation. The epithelium as a whole has increased greatly in height, consisting of from nine to twelve cell layers, and, beneath its surface layer, now displays a well developed, strongly acidophilic stratum corneum and basophilic stratum granulosum. By the end of Stage One a complete detachment of the surface layer has ensued, a detachment which often occurs in sheets and which leaves the stratum corneum now exposed as the lining layer of the vaginal lumen. The first appearance of the cornified layer actually beneath the surface of the epithelium is a remarkable liistogenetic process noted long ago hut not accurately described by Retterer and apparently lost sight of entirely by sub- sequent investigators. The French observer recognized the process in the guinea pig, dog, cat, and rabbit, and took care to note that in some cases cornification did not result, the essential fact being that, in the squamous transformation which ensues, the actual surface cells are exempted. Subsequent to our obser- vations on the rat we availed ourselves of the opportunity to restudy the guinea pig, inasmuch as this feature had been overlooked by Stockard and Papani- colaou, whose model work marks so significant a step in our knowledge of this subject. The same process occurs in the guinea pig. Corner and Pelkan in this laboratory have also been able to confirm its occurrence in the rabbit. Stage Two.—Stage Two, which it will be remembered is characterized by a vaginal smear in which cornified cells first appear, continues to show a high, stratified squamous epithelium (figs. 23 and 24, pi. Ill), which has its stratum corneum at the surface, dehiscence from which furnishes the cells of the smear. The well developed, desiccated stratum corneum aided by the considerable height of the epithelium doubtless causes the dry and lusterless appearance of the mucosa in the gross. The epithelium begins to be noticeably reduced in height by splitting and shedding of the stratum corneum which, after about twelve hours, takes place en masse. The epithelium is still free from leucocytes. Stage Three.—Stage Three is characterized by the continued reduction of the vaginal epithelium, which is now from five to nine cell layers in depth, brought about chiefly through the complete detachment of the stratum corneum and stratum granulosum. This is the stage of accumulation of considerable cheesy or granular masses in the vaginal lumen and the masses consist exclu- sively of the cornified cells. Leucocytes have not yet entered the epithelium in any numbers. Stage Four.—Stage Four, the vaginal smear of which is characterized by the appearance of leucocytes among the cornified cells, exhibits an extensive leucocytic infiltration of the vaginal epithelium which readily accounts for their 24 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA great numbers in the vaginal lumen (figs. 33 and 34, pi. IV). Few or no vestiges of the stratum corneum or granulosum can be found and nucleated squamous epi- thelial cells lie next the surface. The epithelium becomes progressively lower, being usually from four to eight cell layers in height toward the end of this stage. Dioestrous Interval. B. Changes in the Uterus During tlie dioestrous interval the uterus is always slender (fig. 8, pi. I) and with a slitlike lumen, seldom exceeding 2.5 millimeters in diameter (varying from 1.6 to 2.5 millimeters). It is lined by a simple columnar epithelium (fig. 9, pi. V), which possesses next the lumen a delicate cuticular membrane. The character of this epithelium does not undergo any significant change during the early part of the pro-oestrum (fig. 20, pi. Y). Stage One.—No portion of the reproductive system shows so marked a re- sponse to the oestrous wave or one which can be more readily appreciated with the naked eye than does the uterus (see table 1, p. 42). Vascular engorgement, which expresses itself in the turgescence of the vaginal aperture, is also very evident in the uterus of animals killed in the pro-oestral stages (Stage One) ; toward the latter part of this stage there takes place an accumulation of clear fluid in the uterus which distends it to unusual proportions, to dimensions of 5 millimeters in diameter (figs. 16-19, pi. VIII). Whether or not the secretion of this fluid is comparable to the great formation of mucus which Stockard de- scribes as characterizing the uterus of the guinea pig remains to be determined, but it is not unlikely that we have here analogous processes, the fluid in the guinea pig being different in consistency (mucus) and being constantly drained into the vagina, whereas in the rat it is a clear, watery, non-coagulable substance which distends the two uterine cornua tightly, the vagina remaining perfectly dry. This, indeed, constitutes one of the most striking differences between these two rodent forms. The accumulation of fluid within the uterus and the ensuing distention converts the columnar epithelium into a cuboidal one, although its height is not greatly altered. During this time there is no evidence of epithelial impairment and leucocytes are usually absent from the epithelium. Although this oestral distention of the uterus by fluid is unique among mam- mals at present known to us, the production of fluid by the uterus during the time of hvperaemia is to be looked for in other forms. It is perhaps idle to speculate concerning the value of the retention of the fluid within the uterus of the rat, but it is impossible not to suggest that it may serve as a medium through which spermatozoa may swim rapidly toward the oviduct. If spermatozoa are immediately removed from the vasa deferentia of a male rat they will continue for a considerable time to exhibit motility when added to the fluid obtained LONG-EVANS: OESTROUS CYCLE IN THE RAT 25 from the distended uterus at Stage One. It is a fact that enormous numbers of active sperm are found in the uterine fluid in an animal sacrificed a few hours after copulation. Stage Ttvo.—In the early part of Stage Two the uterus exhibits its most marked vascular congestion and reaches its greatest distention (fig. 25, pi. VIII); the distention disappears about the end of the stage. It will thus be seen that this behavior on the part of the uterus parallels closely the exhibition of oestrus on the part of the animal, which, as we have previously stated, may begin with the latter part of Stage One and is usually most evident in the early part of Stage Two. The uterine distention appears to subside rapidly. It is almost never encountered in Stage Three. In spite of this fact, we were at first inclined to look upon resorption as the probable mechanism for its loss because of the per- sistent dryness of the vagina during the whole of Stages Two and Three. We did not, however, reckon with the possibility that the discharge of uterine fluid may occupy but a very short time and that, if it be released suddenly, it could quickly drain through the vagina to the exterior and leave no evidence of its passage in this way unless animals were watched closely. For other reasons this duty eventually fell to our lot, a number of animals being under continuous observation at three-hour intervals, when by good fortune we actually observed the sudden appearance of fluid and leucocytes, most probably of uterine origin, in the vagina of an individual in Stage Two where the vaginal mucosa had been dry for six hours and resumed its characteristic dry appearance again three hours later. Toward the end of Stage Two, when the uterine fluid has disappeared and the walls are flaccid, the normal character of the uterine epithelium reestablishes itself, and the cuboidal type (fig. 26, pi. V) found at the time of maximum dis- tention is succeeded by a columnar epithelium (fig. 27, pi. V). There may now begin a characteristic degeneration of the epithelium, which we have called vacuolar degeneration (fig. 27), because of the appearance and enlargement of cytoplasmic vacuoles. Leucocytes can often be detected in beginning their pene- tration of the uterine epithelium in Stage Two. Stage Three.—In Stage Three, the stage of abundant cornified cells in the vaginal smear, vacuolar degeneration of the uterine epithelium is most typical (figs. 29, pi. VIII, and fig. 30, pi. V). Thus it is possible to say that the time when epithelial dehiscence and degeneration are most expressed in the vagina is the time also of epithelial impairment in the uterus. As we shall establish farther on, this degeneration can be detected in its incipiency before the occur- rence of ovulation, that is, in the time interval between copulation and ovulation. In no cases have we seen degeneration proceed so far as to cause a denudation of the uterus. 26 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Stage Four.—In Stage Four, characterized by the abundance of leucocytes which have extensively infiltrated the vaginal mucosa and have now reached the vaginal lumen, leucocytic infiltration of the uterus is also very abundant. Vac- uolar degeneration reaches its maximum expression in this stage (fig. 36, pi. V). Epithelial regeneration and replacement in the uterus is apparently going on pari passu with the degeneration, and early in the dioestrous interval a perfectly normal, columnar structure is always found (fig. 9, pi. V). From the foregoing it will he clear that the uterus participates in a marked degree in the active changes which characterize all portions of the reproductive system in the various phases of the oestrous cycle, that besides its peculiar secre- tory activity during the time of oestrus proper it undergoes to some extent the same wave of epithelial degeneration and leucoc}dic invasion that characterizes the vagina. It is doubtful whether these are analogous to menstruation, but in the rat, if we were to judge from the time relations alone, we should have no hesitation in deciding between the contrasting theories for the significance of menstruation according to which the uterine changes are considered either as a “freshening” or preparation, for implantation, or as an abortion due to failure of the latter. The chronology of events in the rat would lead us to adopt the first of these two views, for epithelial impairment is not only far advanced by the advent of ovulation but regeneration is complete before the eggs can reach the uterus. The tubal journey in the rat consumes three days. As a matter of fact, even in other forms where uterine degeneration follows rather than pre- cedes ovulation, the egg may consume a time interval in the tubal passage suffi- cient for both uterine epithelial decay and reestablishment. It will probably prove to be the case that in all the Mammalia ovulation delivers eggs into a healthy uterine epithelium. In the Eutheria this epithelium has been freshly regenerated; in the Marsupials it is somewhat older. C. Changes in the Oviduct There is some evidence that the oviduct also participates by structural changes in the events characterizing the oestrous cycle. We shall, however, at this time mention only the most marked and macroscopically detectable change in it, a change which is correlated with its functional capacity for receiving and conducting the egg toward the uterus immediately after ovulation. Although this somewhat anticipates the remarks we shall make under the section devoted to the ovary and ovulation, we may state at once that, in contrast to the condition before ovulation (figs. 38, p. 27, and 40, pi. I), for a period of at least approxi- mately twelve hours after the estimated time of ovulation, the distal folds of the oviduct are distended with fluid (fig. 39, p. 28). These fluid-containing, distal loops of the oviduct also always contain the eggs, the presence of which LONG-EYANS: OESTROUS CYCLE IN THE RAT 27 may be detected in the gross under the binocular microscope whether in the living animal or when the organs are placed in Locke’s or Ringer’s solution, by the movable, opaque mass seen through the rather transparent tubal walls. Sections (fig. 41, pi. I) show that the masses in question consist of the eggs sur- rounded by granulosa cells. In the large majority of normal cases, eggs will thus be found in the fluid-filled, distal fold of the oviduct at the beginning of Stage Four, characterized by the first appearance of leucocytes among the corn- ified cells in the vaginal smear. As the egg enters the intermediate and proximal portions of the oviduct, a similar distention by fluid does not accompany it. Fig. 38. The oviduct at any time when eggs are not present in the distal portion, which is stippled. The part shown by dotted lines is the ciliated ‘ ‘ fimbriated ’ ’ end. The periovarial membrane clings closely to the ovary. X ca. 10. D. Changes in the Ovary 1. OVULATION There can be no donbt of the coordination of ovarian changes with those already described for the vagina and uterus. All that series of changes which culminates in ovulation and the production of corpora lutea has a very definite relation to the various phases of the oestrous cycle as exhibited by the vaginal smear. The exact time relations which ovulation exhibits both with respect to the vaginal epithelial dehiscence and to copulation are, of course, of peculiar importance, for this may have a bearing both on the theories for the significance of menstruation and the fact of occasional unequal embryonic development in pregnancies of the same copulation age. 28 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA The structure of the ovary of the rat at the height of its sexual life and when isolated from males for two or three months is so complex that it would almost seem to forbid analysis. This is due not only to the fact that ovulation, and, consequently, corpora lutea occur at every one of the rapidly succeeding oestrous cycles of four or five days, but also to a remarkable persistence of the corpora, so that many “sets” of them are found in the ovary at any one time, sets which Fig. 39. The oviduct soon after ovulation when eggs are present in the large distal fold shown at the right. The corresponding parts in figures 38 and 39 are stippled. The periovarial membrane is also distended with fluid and separated from the ovary. The “fimbriated” end is in dotted outline. exhibit such slight variations in degree of regression as determined by their structure that it is practically impossible, without some way of marking them, for us to gain any sound notion of their age. It is not unusual to encounter more than fifty of these structures in a satisfactory state of preservation in a single ovary. In addition to this multiplicity of corpora lutea, one finds also in the rat’s ovary a varying amount of follicular atresia, with which the for- mation of interstitial tissue is well known to be connected. On the other hand, LONG-EVANS: OESTROUS CYCLE IN THE RAT 29 the sizes and structure of the normal follicles, though not suffering such great variation, as a matter of fact are related clearly to the oestrous cycle only when marked growth changes take place, so that it might be equally difficult to pre- dict from the follicles with wrhat phase of the oestrous cycle we have to do. The above remarks will serve to indicate that the structure of the ovary does not show marked differences during most of the period occupied by the succes- sion of normal ovulation cycles, but, as we will show in a moment, this is due to the fact that the outspoken changes identified with ovulation occur only for a brief time preceding and succeeding this event. In other words, the growth and rupture of the follicles and the formation of new corpora lutea are phe- nomena which do not occupy any considerable span of time. This fact empha- sizes again the ease with which ovulation could be entirely overlooked did one not have the opportunity of examining the ovary with exact reference to the epithelial changes in the vagina. In order to depict the ovarian changes that can be associated with the oestrous cycle, we may begin with the period in which the follicles are smallest; this is the period immediately succeeding an ovulation. We mav state at once that ovulation occurs during the last hours of the cornified cell stage hSfages-Two and Three), and has always taken place bv the end of that stage, that is, when leucocytes first appear in the vaginal smear. We have previously stated this fact in another way under the caption ‘‘ Changes in the Oviduct” when we said that at the end of Stage Three one may always find eggs from a recent ovulation in the distal folds of the oviduct. At the same time “new” corpora lutea of ovulation are present in the ovary. With the approach of oestrus and during Stage Two, with the consequent hyperaemia of the entire sexual apparatus, growth of certain of the follicles becomes more marked, a process which slowly continues until the more rapid ripening changes take place late in Stage Two. After Stage Three is clearly expressed, the stage in which considerable macroscopic cheesy accumulations of cornified cells occur within the vagina, the steps introductory to maturation may be detected. These consist in beginning indipping of the theca interna folliculi and its capillaries so that the membrana propria of the granulosa is sharply indented at these points. Corner has called attention to this in the sow. The egg as yet is still in the germinative vesicle stage; and the total dimension of the follicle does not exceed eight-tenths of a millimeter (table 2, p. 31). In our experience these changes take place about eighteen hours after the first appearance of cornified cells in the vaginal smear, at the time consequently when the majority of animals are no longer in heat. At some, time during the next few hours, although the follicles do not increase ap- preciably in size, the cumulus displays a distinct corona radiata and the first maturation spindle is formed. The slight thecal indippings noted above may he 30 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA somewhat more marked. There then takes place a final swelling of the Graafian follicles, so considerable that they may attain diameters of nine-tenths of a millimeter. At this time the infoldings of the theca folliculi have become very conspicuous (figs. 42 and 43, pi. XI) and some of the outermost granulosa cells have minute lipoid deposits within them. Now the first polar body may he seen, the second maturation spindle being in place; and in this form the ovum is shed to begin its journey in the tube. While our data indicate that the above statements represent the average condition, they also disclose a considerable variation in the exact time of inci- dence of any of these phenomena. It is true, as we have indicated, that in most animals ripening changes begin about the eighteenth hour after the first ap- pearance of cornified cells in the vaginal smear and that ovulation may he expected at the twenty-fourth hour dating from the same event. However, ovulation may occur at any time in the twelve-hour interval embraced from the eighteenth to the thirtieth hour after Stage One. As regards the exact time involved in the various steps of ripening, we have not been able to establish a reliable opinion. For this it would he necessary, in view of the variation which exists, to accumulate a very large body of data. It is our impression that the remainder of the ripening changes take place very rapidly after the first matu- ration spindle has once been established. The sequence of events in the ripening process which we have been able to discover has some value inasmuch as various criteria have been advanced from time to time from the work of Bischoft on- ward which would enable us to recognize a fully mature ovarian egg. But it is possible that the exact sequence of events which we have described for the rat will not have general application. To attempt a statement of most general application, it would appear that it is not, indeed, until the formation of the first polar body that we may state that the Graafian follicle has nearly reached the time of rupture. Our records of the histological findings in about seventy-five animals form the basis of the above account. Thirty-seven of these animals were sacrificed solely with reference to this question, being killed at intervals of twelve, eight- een, twenty-four, thirty-three, and thirty-six hours after the incidence of corni- fied cells in the vaginal smear. Exact findings are detailed in table 2, page 31. Many of the cases in this table permit us to see the changes which are im- mediately consequent upon ovulation, and, in particular, the inception of the corpus lutemn. The young corpora have very large cavities within them (figs. 41, pi. I, and 44, pi. XI). The closure of the pore of rupture must be very rapid since we have failed to detect it in this relatively large series of cases. Some- what later stages in the formation of the corpus are seen in the cases of animals 3827 and 4093, where the uninvaded central space is smaller. LONG-EVANS: OESTROUS CYCLE IN THE RAT 31 Summary of findings in rats killed for the determination of the time of occurrence of ovulation. (Not included in Table 30) Table 2. Rat Stage Killed hours after first cornified cells Diameter of follicles in millimeters Inpushings of follicle wall by bloodvessels Egg Oestrus 4007 1 0 .65, .70 Some Germinative vesicle 4326 2-3 18 .68, .70 None Germinative vesicle Not yet in heat 3954 2-3 18 + .65, .68, .70, .70 Beginning Germinative vesicle Past 4058* 2-3 18 .80, .83 Present Germinative vesicle 4099* 3 18 .78 Beginning Germinative vesicle 4237 2-3 24 .65, .70, .75 Several Germinative vesicle 4060 2-3 27 + .80, .85 Some Germinative vesicle First time when killed 3994 2-3 36 .70, .75 Beginning Germinative vesicle 4009 2-3 36 Some Germinative vesicle Still in heat 4001 ? 162 .75, .75, .80, .80 Marked! Germinative vesicle Still in heat. Never free of leucocytes. 3824* 2-3 18 .70, .70, .80 Slight Germinative vesicle and first spindle 3815 2-3 15 .75 Present First spindle 4065 2-3 24 JO bo o Slight First spindle 4081* 2-3 24 .65, .75, .75, .90 Some First spindle In heat 4234 2-3 24 .70, .70, .70, .73, .80 Some First spindle 4027 2-3 30 .70, .70, .75, .80 ? First spindle In heat 4036 2-3 . 27 .75, .80, .85 Present First polar cell Still in heat 3784* 2? 30 .80, .80, .88, .90, .90 Large First polar cell Corporalutea 4078* 2-3 18 Large cavity In distal fold 4102* 3 18 Large cavity In distal fold 4070 2-3 21 Large cavity In distal fold + Past 4238 2-3 21 Large cavity In distal fold J 4095* 3 24 Large cavity In distal fold ° Past 4052* 2-3 24 Large cavity In distal fold Past 4056* 2-3 24 Large cavity In distal fold 3799* 3 30 Large cavity In distal fold ■> Past 4008 2-3 30 Large cavity In distal fold £ 4128* 3 30 Large cavity In distal fold ° 4239 2-3 33 In distal fold "5 Still in heat 4033 2-3 33 Small cavity In distal fold Past 4126* 3 36 In distal fold | 4038 4 114 In distal fold *o Past 3827 4 84 Small cavity Near intermediate Past portion 4047* 2-3 30 Small cavity Entering inter- Past mediate portion 4093* 2-3 18 Small cavity In intermediate portion 4066 2-3 30-33 Near intermediate Still in heat portion *Observations started in Stage 1 and samples taken carefully to avoid any possible unfavorable effect on cycle. 32 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Another feature characteristic of the period of ovulation never before to our knowledge mentioned in the literature is the accumulation of fluid in the perio- varial space, cansing a distention of the periovarial membrane (fig. 39, p. 28), which, in pathological cases, may be very great, As will presently be shown, the distal portion of the oviduct also becomes distended with fluid. The fluid in both is probably directly secreted into the cavity in question and is not derived from the follicles, for the fluid in the latter behaves differently in that it alone in fixed preparations produces with the reagents a granular precipitate. The migration of the ovum can be followed with some accuracy. As we have just indicated, in the rat the first polar body is formed by an ovarian egg, as has been demonstrated to be the case for the mouse, guinea pig, rabbit, cat, sow, bat, armadillo, opossum, and Dasyurus. The second maturation spindle for the second polar body must form very soon after the first, for it is in place in the ovarian egg just before rupture, but the second polar body appears not to actually form in unfertilized eggs. During the first twelve hours after ovu- lation, the egg with the second maturation spindle in position moves slowly through the first two or three distended distal loops of the oviduct. During the next twenty-four hours it traverses the intermediate part of the proximal loops, and during the third day it will be found at various positions in the proxi- mal loops, where, in case it has not been fertilized, it degenerates. This degen- eration consists in a characteristic, unequal, direct fragmentation, or segmen- tation, of the protoplasmic mass which may or may not carry with it unequal sized nuclei, as shown by some unpublished observations by Margaret Mann. One may see a similar form of degeneration of the ovarian egg in the case of follicular atresia, where also one may have a rather orderly imitation of blasto- genesis, or parthenogenesis, as L. Loeb has called it. The same thing may also occur in the case of an ovum retained in a corpus (fig. 45, pi. XI). 2. RELATION BETWEEN OESTRUS AND OVULATION The relation between oestrus and ovulation has a very special interest from the standpoint of the time of occurrence of fertilization. We therefore discuss it here and begin with a more accurate description of the time of appearance of oestrus and the variations in that time. Usually oestrus is manifested for a period of from nine to twelve hours, beginning in the last part of Stage One and occupying most of Stage Two. Instances of the long duration of oestrus, of its early occurrence, or of its late occurrence are not unusual; nor must we fail to note another peculiarity, that is, the occasional occurrence of the manifestation of oestrus for only three hours or a similar very short interval. We have accumulated these results by the method already indicated, i.e., by offering individual females to a cage of males at the three-hour intervals when a record was kept of the vaginal smears. LONG-EVANS: OESTROUS CYCLE IN THE RAT 33 Early and late exhibitions of oestrus have a special significance when brought into relation with ovulation. They might enable us to explain the variation found in the development of young embryos when the only age criterion is the copulation date. Table 2 indicates that ovulation may occur as early as eighteen and as late as thirty hours after the first appearance of cornified cells in the vaginal smear. Oestrus may be exhibited as early as three hours before the appearance of cornified cells or as late as twenty hours after. Did ovulation succeed oestrus by an invariable interval of time, the phenomenon of unequally developed embryos at the same copulation age would have to be explained by the unequal rate of ascent by the spermatozoa through the genital passages. This explanation suffers, however, first, from the belief in the limited viability of unfertilized eggs, and, secondly, from the fact that the enormous number of sperm ejaculated make it likely that the eggs are reached in an approximately uniform time. The last objection gains weight in the case of those forms which, like the rat, have ovulation normally at a very considerable time interval after copula- tion. In the case of the rat almost a whole day separates these events. Long has demonstrated that sperm may reach the distal part of the oviduct in the mouse, a related form, within .four hours. In view of all the facts, then, it is extremely probable that, in the rat, sperm will usually await the arrival of the ovum for a considerable interval of time, i.e., fertilization ensues immediately upon ovulation. Now it is very likely that ovulation does not necessarily follow oestrus in a uniform period of time, for there is reason for relating it to the general progress of the oestrous changes taken as a whole, an indication of the orderly progression of which we have found in the vaginal smear. In the case of very late exhibition of oestrus associated with precocious ovulation there might, then, be an actual transposition of these two events and a failure of fertilization through the inability of ova to continue to maintain their vitality until reached by sperm. It is evident that the cases of speediest fertilization following copulation will occur under conditions in which oestrus and ovulation are nearest together, providing sufficient time (at least four hours) elapses for the sperm to accomplish the uterine and tubal journey; and, conversely, the greatest delay in fertilization would occur in those instances in which oestrus and ovulation are separated as far as possible from each other. For example, if oestrus is manifested three hours before the occurrence of Stage One and ovulation thirty hours subsequent thereto, the sperm would have to wait thirty- three hours before fertilizing the eggs. Thus age differences in embryos dated from the hour of copulation could easily arise and could conceivably involve somewhat over a dav. 34 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA 3. THE POST-PARTUM OVULATION In the rat, as is known in other animals, an ovulation occurs shortly after parturition. The post-partum ovulation may take place at any time between the sixteenth and twenty-fourth hours after littering, as indicated in table 3. Table 3. Showing number of hours after parturition when eggs were procured from the oviduct. Hours post partum No. of instances when eggs were found No. of instances when eggs were not found 15 1 16 1 17 3 1 18 5 1 19 5 2 20 3 21 5 2 22 5 5 23 2 24 11 6 25 7 26 4 27 1 28 2 29 3 30 3 32 3 33 2 36 1 37 1 43 1 44 1 69 18 4. CORPORA LUTEA Whatever may be the exact mechanism of follicular rupture, we must regard this as an influence widely distributed in the ovary as a whole, for ovulation is characterized by the practically simultaneous rupture of all the follicles be- longing to that particular “set.” In the examination of scores of complete series of ovaries at times just preceding and subsequent to ovulation it is re- markable how seldom one encounters instances of a lack of synchronism in this phenomenon. At the same time that we make these statements, we would call attention to the occasional apparent failure of a single, or, at most, of two follicles to undergo rupture along with their mates, and to the occurrence of follicular atresia in such large follicles, but whether or not the atretic process has begun before ovulation we are unable to state. LONG-EVANS: OESTROUS CYCLE IN THE RAT 35 Corpora lutea atretica sensu strictu. We have also encountered a rarer condition: a highly peculiar type of atresia in those occasional follicles where there has apparently been no rupture. In the cases to which we refer the granulosa does not undergo the degeneration characteristic of atresia but proceeds to the formation of lutein cells similar in all respects to those of the fresh corpora produced from the other ruptured follicles of this set, but surrounded by theca cells which remain in place and are transformed into interstitial tissue. Corpora lutea with retained ova. Finally, there is the somewhat commoner process, namely, the production of typical corpora lutea by follicles in which the eggs hav'e still been retained in situ (figs. 45, pi. XI, and 59, pi. YI). In such instances the thecal cells behave in no way different from their role in the formation of normal corpora lutea. They do not produce interstitial tissue, but to an extent not yet determined they partici- pate in the ingrowth of connective tissue which accompanies the blood vessels of the corpus. In other words, a typical corpus luteum is formed, in the center of which, however, the egg is found. This phenomenon is by no means unknown, but we doubt whether the frequency of its occurrence has been suspected by any previous investigator. In order to obtain more data upon this point, we undertook to examine carefully all those cases in which the number of eggs discharged at an ovulation could be ascertained with certainty as well as the number of corpora lutea from that particular ovulation (table 4). In these cases, ova in a good state of preservation were encountered in the distal portion of the oviduct and their number could be counted with reliability. Thirty- seven such cases were found suitable for our purposes. In each case we studied only a single oviduct and ovary. In twenty-five of them the number of ova encountered corresponded strictly with the number of fresh corpora found in the corresponding ovary, and in twelve cases the number of corpora was in excess of the number of eggs found in the oviduct. In some of the latter in- stances, moreover, it was possible to discover that the ova which were missing in the oviduct were at the center of certain of the corpora; but it is to be pointed out that a rather rapid degeneration of these retained ova takes place, so that in our opinion in many instances where they previously existed they were not encountered by us. It would appear, then, that about one-third of all ovulations exhibit one or more instances of retention of the ovum within the corpus luteum, but the total number of normal and abnormal corpora which we have thus assumed to exist in these thirty-seven ovulations stand in the ratio of nine to one. Since, in our experience, about ten follicles mature at each ovulation (table 4), the above statement might imply that one of the corpora in each ovulation is apt to have 36 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA a retained ovum, but the above analysis would indicate that only about one-third of all ovulations show such abnormalities and that the nine to one proportion of typical to atypical corpora is due to the multiple occurrence of abnormal corpora in the same ovulation. Table 4. Table showing correspondence between the number of eggs in the oviduct and the number of corpora formed at that ovulation (in one ovary and oviduct). Average number of corpora lutea in 54 cases = 5.37 in one ovary Average number of eggs in 50 cases = 4.8 in one ovary A r Average number of corpora lutea in 37 selected cases in which both corpora and eggs in oviduct could be counted = 5.4 in one ovary k Average number of eggs in same 37 cases = 4.8 in one ovary B Average number of corpora lutea produced at each ovulation. It would appear that on the average five corpora lutea are produced in each ovary at any one ovulation. One or both ovaries from one hundred and eighty- two ovulations have been analyzed in order to reach this conclusion. The aver- ages obtained are exhibited in the subjoined table. Table showing number of corpora per ovulation in cases in which the determination is certain. 95 pregnant animals Average 4.93 in one ovary 28 nursing animals Average 5.7 in one ovary 40 animals—corpora lutea of first ovulation Average 5.45 in one ovary 19 animals—corpora lutea of second ovulation Average 5.05 in one ovary J General average 5.20 in one ovary"] Table 5. Although five is thus the average number of corpora per ovulation in a single ovary, as many as ten corpora and as few as one have been encountered. The distribution of instances of the number of corpora lutea produced at a single ovulation may be seen from table 5a, which is merely a re-arrangement of the data from which table 5 has been secured, together with the addition of seven new cases. LONG-EVANS: OESTROUS CYCLE IN THE RAT 37 iSize of corpora lutea of ovulation. The corpora lutea of ovulation when fully formed usually measure about 1 millimeter in diameter. The maximum measurements which we have re- corded are of diameters of 1.2 millimeters. Succession of the corpora by an- other set is not immediately indicated by size regression of the first set, whose dimensions are indeed not perceptibly reduced even by the second succeeding ovulation when their greatest diameter is still about a millimeter. As far as we have been able to determine, the corpora lutea of ovulation do not differ from those of pregnancy or pseudopregnancy except in the larger size which the latter type of corpora usually attain. The corpora of pregnancy, however, do not exceed those of ovulation until after the tenth day of gestation. Table 5a Table showing the number of instances in which various numbers of corpora lutea were pro- duced in a single ovulation. Number of corpora lutea produced at a single ovulation Instances 1 3 2 8 3 21 4 42 5 41 6 29 7 26 8 4 9 2 10 3 Functional life span of corpora lutea ovulationis Two points of view make it important for us to attempt to determine with some precision the functional life span of the corpora lutea of ovulation. We refer to the exact relation of the corpora lutea to the oestrous cycle and to the confusion created by the retention of corpora lutea in the rat for many oestrous cycles succeeding their formation. It is obvious from the great number of cor- pora which occur in rat ovaries that any outspoken degeneration of the corpora lutea of any particular ovulation can not occur before the next oestrus brings on the succeeding set. But since in other forms corpora lutea, as long as they are functionally active, are supposed in some way to prevent ovulations and to hold off the oestral degenerative changes which characterize the mucosa of the reproductive system, it was important for us to determine whether or not the conditions in the rat invalidate this assumption. In other words, it was import- ant to detect, if possible, any morphological changes which would indicate a 38 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA diminution or change in the function of the corpora lutea of ovulation just pre- ceding the next oestrus. We have found these morphological changes in size and distribution of the fats which brown with osmic acid, and interpret this change in the amount, or, more properly, aggregation, of the lutein cell fat to be the expression of a change in the physiology of the cell. Our assumption seems the more likely inasmuch as exactly parallel changes occur in the corpora lutea of ovulation, of copulation or pseudopregnancy, of gestation, and of lactation, as we will show later on; and in the latter cases these changes are definitely related to the cessation of the gravid or of the lactating state. In the ripe Graafian follicle just preceding rupture one may observe that the cells of the theca interna are well laden with fair sized lipoid spherules which blacken readily with osmic acid, whereas the granulosa cells, later to become the lutein elements, have no appreciable fat, or, at most, minute brown granules, in their outermost layers next the basement membrane which separates them from the theca interna. Very young corpora have the same minute gran- ules (fig. 47, pi. VII). Thirty hours after the rupture of the follicle, this differ- ence in the fat content of the theca interna and granulosa has disappeared, and the enlarged granulosa elements, now the lutein cells, possess small lipoid drop- lets which brown in the osmium tetroxide and have the dimensions and distri- bution shown in figure 48, plate VII. In order to obtain uniformity in meas- urement, the granules were measured by means of a special eyepiece. In this ocular was placed a disc of glass on which had been sprayed exceedingly minute particles of India ink, all of which were then removed except nine, which were selected to form a graded series. The lipoid granules in the preparation were easily compared with these spots in the eyepiece and designated according to the number of the spot or spots with which they corresponded. The spots are shown in figure 46 as they would appear in a drawing at 750 diameters. Fig. 46. The spots on the ocular micrometer disc and 10 divisions of a millimeter divided into hundredths drawn with the same objective (Zeiss 2 mm. apochromat) and ocular (6 compensating) used in measuring lipoid granules in lutein cells. The spots are numbered from 1 (smallest) to 9 (largest). Larger granules are meas- ured in multiples of 9 (i.e., 3 X 9). X 750. Same magnification in figures of lipoids, plate VII. LONG-EVANS: OESTROUS CYCLE IN THE RAT 39 The lutein lipoid droplets are highly refractive when viewed in a drop of Ringer’s solution in which the fresh lutein cells have been crushed by a cover glass; when fixed with formalin, they are stained red by the application of a saturated aqueous solution of Nile Blue Sulphate to the frozen sections, which are then rinsed in distilled water and transferred to a 1 per cent sodium hy- droxide solution. We have found that the best way of fixing and preserving them is by the employment of Meves’ modification of Benda’s Fluid (Meves and Duesberg, 1908), and by the after treatment of such fixed material with pyroligneous and chromic acid and potassium bichromate. The ovary is transferred from Benda’s fluid after 48 hours, rinsed in distilled water, put into a solution of equal parts of 1 per cent chromic acid and pyroligneous acid for 48 hours, rinsed again in distilled water, and is then transferred to 2 per cent aqueous bichromate of potassium for 48 hours. After this it is washed for from 12 to 24 hours in distilled water and dehydrated, cleared for from 2 to 4 hours in cedar oil, transferred to xylol and imbedded in paraffine. The tissue may be sectioned at any thickness and the sections mounted on slides; the lipoid may then be examined directly with the oil immersion lens, the oil serving to make the paraffine sufficiently transparent. Without the treatment with pvroligneons and chromic acid and potassium bichromate, the lipoid gran- ules would dissolve easily in the xylol and balsam. Even after this treatment, they are sufficiently soluble in balsam that preparations a year old are some- what reduced in intensity when compared with uncovered paraffine sections. These granules are small and strikingly uniform in size, especially in the same cell, although they are somewhat larger than the extremely minute fat droplets which we have mentioned as appearing in the granules at about the time of rupture, and are larger than the very small lipoid granules which, we shall show later on, are characteristic of the cells of the corpus luteum of lacta- tion. They are nevertheless very much smaller than the granules of ating corpora, which, moreover, never have the striking uniformity in size. As a rule, they are also uniformly distributed in the cell and even in the cases where that distribution is uneven they are not so crowded together as to be in contact. At the time of onset of the next oestrus some of these lipoid deposits in the corpus are appreciably larger in size and their number is increased (fig. 49, pi. VII). These changes occur first in the inner hemisphere. It is this morpho- logical change that, in our conviction, is correlated with a functional impairment of the lutein cells. The lipoid changes, of course, progress farther; they attain a maximum and regress before the corpus actually begins to diminish in size or to undergo what would ordinarily be called outspoken degeneration. It is im- portant to know that the lipoid changes are very marked before any other de- generation of the lutein elements or diminution in the size of the corpus can be detected. We have already noted that a diminution in size of the corpus does not usually take place until the beginning of the third cycle succeeding a corpus of ovulation. Early in the succeeding oestrous cycle the sizes of the lipoid 40 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA granules continue to be even more clearly unequal and the granules are more irregularly distributed in the cell, while their number is greatly increased. This appearance of larger lipoid granules may continue until the cell may contain one or more bodies considerably larger than the nucleus, but succeeding this phase the lipoid decreases. At the beginning of the third cycle the lipoid becomes much reduced in amount; and it is when this phase of diminution occurs that there is at last perhaps a slight reduction in the size of the corpus as a whole. By the time that the lipoid content of the lutein cells is notably reduced, macrophages are clearly abundant in the corpus and are gorged with fatty substances which, in spite of the after treatment with pyroligneous acid, etc., is never black, but brownish; whereas the Nile Blue Sulphate method shows that some of these great phagocytes have fatty acids within their vacuoles, substances never present in the lutein cells and indicating a possible splitting of the ingested lutein cell fat. Ovaries fixed with osmic acid at any time during the succession of normal ovulation cycles thus show sets of corpora lutea which vary in the size, number, and distribution of their lipoid bodies in accordance with their age in the way which we have just described. (Consult also pi. VII, and figs. 72, 83, 51, pi. VIII; 53, 89, pi! IX; 73, 54, 52, 82, pi. X.) We may summarize our views by stating that our studies on the morphology of its lipoids indicate that the corpus luteum cell has reached its greatest func- tional activity at the end of the oestrous cycle which gave it origin, and that at the beginning of the next oestrus it is in a state of regression. Accordingly, any corpus that contains large lipoid granules is one which is already degenerating. The considerable difficulty in deciding in the case of the corpora lutea of ovu- lation whether degeneration precedes or follows the exact time of incidence of the next oestrus has proved easier to surmount in the case of the corpora of gestation and lactation, where, as we will show later, degenerative lutein changes are separated by a somewhat longer time interval from the next oestrus. The reader may naturally inquire how it is that we have been emboldened to identify, in the complex ovary of the rat, any particular set of corpora lutea of ovulation with a known ovulation; and, furthermore, how we knew we were dealing with the corpora of ovulation toward the end of of an ovulation cycle, inasmuch as these cycles vary in length. It is consequently necessary to explain how we oriented ourselves with certainty in this matter. Lactating mothers were treated with intraperitoneal doses of the vital dye, Dianil Blue 2 R (see p. 60), in order to mark with unmistakable blue deposits the lutein cells of the corpora of lactation and of the preceding pregnancy. The young were then removed and the first sponta- neous oestrus, as determined by the vaginal smear, noted, when the corpora of that ovulation were examined by extirpating one of the ovaries. Possession of the other ovary enabled the animal in most instances to exhibit the next oestrus at a normal time (i.e., four to five days later), and in such cases we could he certain that we were dealing, in the case of the extirpated ovary, with the condition of the corpus at particular times in dioestrous intervals of normal length. A fuller explanation of the matter may be afforded by recounting a typical protocol from many such which were employed in our work. (See description of fig. 53, pi. IX.) LONG-EVANS: OESTROUS CYCLE IN THE RAT 41 Instances can be fonncl of the existence of from fifty to seventy-five corpora lutea in a single ovary of the rat, though many of these are badly degenerated and some diminutive. In all cases the last four or five sets (i.e., from twenty to thirty corpora) are substantial, fairly well preserved structures, a fact which tallies well with our counts of the number of well preserved corpora that are encountered about twenty days postpartum. An analysis of many of the latter cases gave us an average of thirty corpora lutea at twenty days post-partum, so that when five of these are allowed as corpora of pregnancy and five as corpora of the post-partum ovulation, twenty are left to represent, probably, four ovu- lations of five corpora each at five-day intervals. No such persistence of the corpora lutea of ovulation takes place during the gravid condition, for on the twentieth day of pregnancy, instead of encountering thirty corpora in each ovary, as occurs twenty days post-partum, we encounter usually merely the five large corpora lutea of gestation, which have thus not merely inhibited ovulation, but have also brought to decay and complete resolution all other corpora occur- ring in the organ. We were able to make the above observations by taking advantage of the fact that at the end of pregnancy practically no corpora of ovulation are visible. By permitting the corpora of the spontaneous ovulation on the day of littering to occur and by breeding the animal at the incidence of the next oestrus, we secured the conditions required, that is, we were able to say that the only corpora of ovulation present toward the end of the second gestation were those of the first post- partum ovulation which intervened between the two pregnancies. But in order to still further distinguish these with certainty from the corpora of the first pregnancy, we administered a vital dye to the animal from the sixteenth to the twentieth day of the first gestation, thus marking clearly the corpora of the first gestation. There were thus three kinds or sets of corpora in these ovaries and only three—the corpora of the first pregnancy which were blue, those of the post- partum ovulation, badly degenerated, and those of the second pregnancy. E. Correlation of Changes in the Vagina, Uterus, Oviduct, and Ovary The correlation of the changes which we have thus demonstrated to take place during each one of the stages of the oestrous cycle in the various portions of the reproductive system of the rat is best summarized and visualized in table 1. F. Length of the VxVrious Portions of the Oestrous Cycle In order to have a correct conception of the length of the various portions into which, by changes in the vaginal smear, we have divided the oestrous cycle, it was necessary to make observations on a relatively large group of animals separated by the shortest practicable time interval. After some preliminary work with especial reference to the number of animals which had to be handled, we settled upon the proper interval between examinations as one of the duration of three hours and preliminary to this work made for some weeks a daily ex- amination of several hundred animals in order that we might employ for the 42 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA final work those with regularly recurring four to five day cycles. About one hundred animals were thus selected for this work, but the abnormalities intro- duced by this handling did not permit more than about sixty of them to complete their oestrous cycles in a time to be regarded as practically normal and hence available for this purpose. In many cases the observations were carried well into a second cycle so that in the same animal two or more instances of any one portion of the cycle were secured. Our results are best summarized in tabular form. Table 1. Schematic outline of changes in the reproductive organs of the rat during the oestrous cycle. Stage Living Animal Histology of vaginal mucosa Uterus Ovary and Oviduct 1 (12 hrs.) Vaginal mucosa slightly dry. Smear of epithelial cells only. Lips a little swollen. In heat toward end. Many layered (8-12) .08- .1 mm. thick. Mitoses active. Cornified layer under superficial layer of epithelium. No leucocytes. During Stage 1 uterus be- comes distended with fluid increasing in dia- meter from 2.3 to 3.7 mm. Follicles large 2 Vaginal mucosa dry and lusterless. Smear of cornified cells only. Lips swollen. In heat. 7-11 layers of cells .08-. 1 mm. thick. Cornified layer well formed and superficial. No leuco- cytes. Mitoses fewer. Reaches greatest disten- tion (5 mm.) and thin- ness of epithelium and then regresses to dia- meter of 1.8 mm. Vacu- olar degeneration some- times begins. Follicles largest. Eggs may undergo matura- tion. 3 (2 and 3 27 hrs.) As in Stage 2, but corni- fied material abundant (cheesy) and animal not in heat. 5-9 cells thick. .064 mm. thick. Cornified layer loose and finally com- pletely detached. No leucocytes. Mitoses still fewer. Diameter of uterus about 2.0 mm. Epithelium undergoing vacuolar degeneration. Ovulation. Secretion of fluid into periovarial space and oviduct. 4 (6 hrs.) Vaginal mucosa slightly moist. Smear of corni- fied cells and leucocytes. Swelling of lips gone. 4-8 cells thick. .062 mm. thick. Cornified layer gone. Many leucocytes. Mito- ses increasing. Diameter of uterus 2.2 mm. Some vacuolar de- generation but also re- generation. Young corpora lutea. Eggs in oviduct. Fol- licles smallest. 5 Dioestrous interval (57 hrs.) Vaginal mucosa moist, glistening. Smear of leucocytes and epithelial cells. Variable amount of mucus. 4-7 cells thick. .042 mm. thick. Leuco- cocytes. Mitoses not numerous. Diameter 1.7 mm. Epithelium undergoing regeneration. Follicles of various sizes Corpora lutea continue to grow. Eggs travers- ing oviduct throughout early interval. LONG-EVANS: OESTROUS CYCLE IN THE RAT 43 Table 6. Table showing the length in hours of Stage One of the oestrous cycle of the rat. (Vaginal smear consisting of abundant, uniform- sized, nucleated, epithelial cells only.) Length of Stage 1 in hours Number of instances 3 3 6 8 9 11 12 40 15 20 18 17 21 14 24 5 27 1 30 1 Table 7. Table showing length in hours of Stages Two and Three of the oestrous cycle of the rat. (Vaginal smear consisting of cornified cells only.) Length of combined Stages 2 and 3 in hours Number of instances 12 2 15 0 18 3 21 3 24 7 27 14 30 6 33 13 36 6 39 7 42 9 45 8 48 6 51 3 54 4 57 2 60 2 63 3 66 0 69 2 72 0 75 1 78 0 81 1 44 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Table 8. Table showing length in hours of Stage Four in the oestrous cycle of the rat. (Vaginal smear consists of cornified cells and leuco- cytes.) Length of Stage 4 in hours Number of instances 3 23 6 40 9 19 12 13 15 7 18 4 21 0 24 0 27 1 Table 9. Table showing length in hours of Stage Five, or the Dioestrous Interval, in the oestrous cycle of the rat. (Vaginal smear consists of leucocytes and nucleated epithelial cells.) Length of Stage 5 in hours Number of instances 30 2 33 1 36 1 39 1 42 6 45 3 48 10 51 4 54 6 57 8 60 8 63 5 66 7 69 3 72 3 75 5 78 1 Mode: 48 hours Average: 43 hours LONG-EVANS: OESTROUS CYCLE IN THE RAT 45 Table 10. Summary. Table showing length in hours of the component parts of the oestrous cycle of the rat. Stage Mode Average One 12 hrs. 14.2 hrs. Two and Three 27 hrs. 38 hrs. Four 6 hrs. 7.8 hrs. Five 48 hrs. 53 hrs. G. Total Length of Oestrous Cycle Observations were made on somewhat over three hundred females from four to six months of age and accordingly in full sexual vigor, having been from the time of weaning isolated from males. These animals formed the basis for our determination of the oestrous cycle in the rat as lying between four and six days. We believe that the data present the normal spontaneous ovulations in this animal. Influence of the male was obviated by isolation of the sexes. More- over, since on several occasions a female was observed to try to act the part of the male, lest even females might possibly have some influence on one another, twelve animals were kept in solitary confinement for about a month, at the end of which time it was clear from the cycles that there was no difference between the ovulation periods of such animals and those of females allowed to live to- gether in small numbers. Complete data on which our results have been based may be found in Table 37, appendix. The data are also summarized in Table 11. The range of variation is from three days to twenty-eight, two cases being re- corded, even, of spontaneous oestrous cycles of thirty-nine days’ duration. As will be clear from the data, we are not inclined to attribute much importance in an understanding of normal phenomena to any cycle in excess of eight days, for 92 per cent of even a miscellaneous lot of animals have cycles that fall within the eight-day period (average 4.8 days). So great, in fact, is the proportion of the combined number of instances in which cycles of four, five, and six days were observed that we do not hesitate to designate this time interval (average 4.6 days), which comprises 82 per cent of our actual observations as the normal interval for this animal. 46 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Table 11. Table of observed instances of oestrous cycles of various lengths. Length of cycle in days Number of instances 3 65 4 789] [Aver- 5 634 j>82% jage 4.6 92% Average 6 233 ] (days 4.8 days 7 69 8 60 9 30 10 24 11 16 12 22 13 and over 57 1999 General average 5.4 days The reader, however, will desire some explanation for the fairly frequent occurrence of longer cycles and without further work upon this subject we pre- sent several concrete suggestions. Our colony was fed upon the fluctuating ration constituted by table scraps, and we have observed a tendency to the pro- longation of the oestrous cycle in all cases of undernutrition; so, also, some disturbance in the exact regularity of feeding hours and the excitement pro- duced by frequent handlings has given evidence of a slight prolongation of the cycle. Furthermore, impairment of respiration, which is invariably produced by allowing animals to pile up together in large cages and which we believe may even occur where three or four animals are confined narrowly, has, in our ex- perience, lengthened the oestrous cycle. We thus have in our hands an ex- tremely sensitive index of the well-being of the young adult female rat, an index more adequate to portray a sound physiology than the appearance of bodily activity, a glossy coat, normal weight, or any other sign known to us. The short- est oestrous cycle which occurs with any regularity is that of four days, and this remains also the most frequently occurring length of cycle in all our work. We are even, in fact, inclined to regard it as the true normal cycle, explaining the high number of instances of the five-day cycle as due to slight, but constantly operating, inadequacies in the hygiene to which our colony, in spite of our care, was submitted, though it will remain for future work with a superior hygiene to substantiate this surmise. We give our data exactly as they were found in order that the material with which we have dealt may be fully known and may be compared with that of subsequent investigators. LONG-EVANS: OESTROUS CYCLE IN THE RAT 47 It is only necessary, in conclusion, to note that in our experience but slightly more than half of a miscellaneous stock of animals give oestrous cycles of such unfailing regularity that a discrepancy of more than forty-eight hours in the length of any one of them did not occur, and that, furthermore, even in these instances cases of an inexplicably longer cycle interpolated in a long series of regular cycles occurred. We are thus acquainted with the necessity not only for a very superior hygiene hut also for an exact individual oestrous history of every animal upon which reliable data as to the experimental physiology of the sexual system are desired. In the material which this monograph presents no conclu- sions have been drawn except on the basis of the study of animals which we have submitted to this exact control. H. Comparison of the Oestrous Cycle in Rat and Guinea Pig The two forms in which the oestrous cycle is best known are the guinea pig and the rat. In making a brief comparison between them it may he said at the beginning that the correspondence in general between the two forms is close. After observing about twenty-two guinea pigs for several months, we feel that the correspondence may be even closer than indicated by the published de- scriptions of Stockard and Papanicolaou. In the first communications on the oestrous cycle in the rat (Long, 1919; Long and Evans, 1920) four stages were described. They were numbered from 0 to 3, instead of 1 to 4, solely for the purpose of making clear the homologies with the guinea pig, for it wTas believed that in the rat Stage 0 was an earlier stage than Stage 1 in the guinea pig. Furthermore, the rat has no stage corresponding to Stage 4 of the guinea pig. In other words, it seemed as though the oestrous changes in the rat began and ceased earlier than in the guinea pig. In their latest paper (19.19) Stockard and Papanicolaou have divided their Stage One into two periods, thus making the first period of Stage One correspond to our old Stage 0 and the second period to our old Stage One. In this paper the stages are renumbered One to Four, instead of 0 to Three, since we believe that we have established an accurate foundation or standard for both rodents and one which we hope might serve for other forms of mammals. To make the guinea pig conform and also simplify it we would suggest that the two periods of Stage One be called Stages One and Two. and the former Stages Two to Four be designated Stages Three to Five. The older and the new terminologies of both rat and guinea pig would then be indicated in the following table: 48 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Numerical designation of stages in oestrous cycle of rat and guinea pig Rat Old Rat New Guinea pig New Guinea pig Old 0 1 1 1st period of Stage 1 1 2 2 2d period of Stage 1 2 3 3 2 3 4 4 3 5 4 In the following comparison between the rat and. the guinea pig the stages will be designated as in the two middle columns of the above table. Stage One (first period of old Stage One of guinea pig).—In both animals there is a dry condition of the vaginal mucosa not very evident in either form. In the smear there are no leucocytes, but only squamous epithelial cells. In the rat these epithelial cells are somewhat different from those of the interval in that they are rounder and contain small granules or vacuole-like bodies which are much clearer in fresh preparations than in stained. In the guinea pig we have also been able to recognize these distinctive characteristics, so that with experience one can distinguish between these epithelial cells and those of the interval or of Stage Three of the guinea pig. A condition in the rat not yet described in any other mammal is to be found in the secretion of fluid into the uterus, which causes great distention of the latter. This accumulation of fluid reaches its maximum at the end of the stage and into the next stage (Two), when it disappears probably by escape through the vagina. It may be that this fluid corresponds to the mucous secretion in the guinea pig, which, as pointed out in the next paragraph, also escapes into the vagina. Copulation may occur late in this stage in the rat, not at all in the guinea pig. Stage Two (second period of old Stage One of guinea pig).—In the rat the vaginal mucosa is conspicuously dry, having a strong, superficial cornified layer from which a few elements become detached to form the typical cornified cells of the smear. This is the usual time of copulation in both forms. In the guinea pig Stockard and Papanicolaou (1917) describe a transition to the next stage (old Stage Two) marked by the presence of a few of these cornified cells. We have found them in the guinea pig in as large numbers as in the rat, the ap- pearance being almost identical, coming after the superficial epithelial cells. Moreover, in a few instances the vagina was dry at first, to become flooded with mucous almost immediately. We believe that Stockard and Papanicolaou have underestimated the importance of this condition, representing as it does the cornified layer in the rat. We feel assured of this not only because the presence LONG-EVANS: OESTROUS CYCLE IN THE RAT 49 of such a layer was seen by Retterer (1892) but because we have seen it in sec- tions from a guinea pig which we killed at this period, in which we also saw unmistakable evidence of the occurrence of the superficial epithelial layer which is the source of the epithelial cells of Stage One in the rat and probably also in the guinea pig. Stockard and Papanicolaou (1919, pp. 234-235) evidently saw this cornified layer, but interpreted it as the whole “epithelium separated from the underlying tissue. ’ ’ In several instances we also have found this layer shed en masse and preserving the form of the lumen of the vagina, and under the microscope satisfied ourselves that it consisted chiefly of a continuous layer of cornified cells, to which, however, there adhered large numbers of the epithelial cells characteristic of the next stage of the guinea pig, Stage 3 (old Stage Two). Whereas in the rat Stage Two is dry, in the guinea pig it is characterized an abundance of mucous derived from the uterus. Stage Three (old Stage Tivo of the guinea pig).—This stage is marked in both animals by the desquamation of large numbers of cellular elements to form a “cheesy mass” which in the rat consists of non-nucleated cornified cells de- rived from the cornified layer and probably the continued cornification of the next deeper, flattened stratified layers. In the guinea pig this 11 cheesy” material is made up of nucleated epithelial cells. In this animal the vaginal epithelium is very thick (more so than in the rat) and may well be considered as giving rise to the cheesy mass, after losing its stratum corneum en masse, by the rapid ex- foliation without cornification of the deeper layers of cells until it later reaches the condition shown in their figure 17 (pi. 5, 1917). In both animals leucocytes collect under the epithelia of both uterus and vagina. It is in this stage, also, in both animals that ovulation occurs, perhaps a little earlier in the rat than in the guinea pig. Stage Four (old Stage Three of guinea pig).—This stage is characterized in both forms by a sudden wave of leucocytic invasion of the mucosa in both vagina and uterus. In both there is continued destruction of these epithelia. In the rat the‘vaginal smear now consists of cornified cells and leucocytes, in the guinea pig of the corresponding epithelial cells and leucocytes. Destruction of the uterine epithelium in the rat goes on by a sort of vacuolar degeneration resulting in the loss of single cells rather than in the loss of large areas as in the guinea pig. This degeneration may, however, begin earlier in the rat. Stage Five (old Stage Four of the guinea pig).—This stage is not repre- sented in the rat by the appearance of blood externally, as in the guinea pig. In both cases regeneration begins, but whereas in the rat the vaginal epithelium is renewed from the deep layer which is never lost, and the uterine by mitosis of those cells remaining, in the guinea pig the vaginal mucosa is said to be regen- erated from the portions in the deep folds, and the uterine from the uterine glands. 50 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA The interval is much longer in the guinea pig, both animals are quiescent during it, and in both the vaginal smear consists of epithelial cells and leucocytes. It will be seen, then, that in spite of several divergencies in detail, there are the same fundamental changes in all the reproductive organs of both rat and guinea pig and the same correlations between these organs. VI. THE ATTAINMENT OF SEXUAL MATURITY A. Age at First Copulation Sexual maturity must be defined as the earliest stage at which the animal is capable of producing young. Satisfactory evidence as to the advent of sexual maturity might be supposed to come from a knowledge of the date at which first litters are born in an animal colony in which the sexes are reared together. Under these circumstances it is unlikely that the first oestrus will be experienced by a female who does not have an opportunity to copulate. We present data upon this point in the subjoined curve A, page 53, which shows the earliest copu- lation known in the case of one hundred ninety-nine individuals. In some of these cases the date of copulation was calculated from the date at which the first litter was born, in other cases we had the opportunity, in our round of daily examinations, to observe the vaginal plug. It will be noted that most instances of the first known sexual congress occur in rats between sixty-three and one hundred and seven days of age; taking the entire data, the average age at the first coitus is 92.7 days. It is apparent, however, that these data are neither necessary nor adequate to answer the question as to when ovulation begins in the rat, for it is conceiv- able that there could be some lack of correlation in the development of the ovaries and vagina or that typical oestrous desire is not sufficiently manifested at the inception of the first cycle to lead to mating. B. Age at Establishment of Vaginal Orifice and at First Oestrous Changes in Vaginal Smear We have consequently been led to examine with some care the time of estab- lishment of the vaginal orifice and to correlate it so far as possible with the condition found in the ovaries. In immature animals the lumen of the vagina does not extend to the exterior, but is closed by what appears to be a thick mem- brane which comes to resemble a cicatrix by its glistening character and appears to become gradually thinner and thus ruptures. This rupture would appear to occur by the widening of a minute aperture centrally located in the vaginal membrane, but it often takes place in two minute apertures bilaterally placed, leaving a thin median cord of epithelial tissue extending across the vaginal orifice dorsoventrally. We have seen this cord persist until the first parturition. LONG-EVANS: OESTROUS CYCLE IN THE RAT 51 In a large proportion of all normal animals (in our estimation in a little less than one-half) the actual breakdown of the vaginal membrane is correlated with the first oestrus and is heralded by a swelling of the lips of the incipient opening, just as is the case with the onset of Stage One in the oestrous cycles of the adult. In all these cases at the time of occurrence of the actual opening a sampling spatula withdraws a vaginal smear showing either Stage One or Two, i.e., it will indicate that the process of cornification has occurred and either the superficial, characteristic, small, nucleated cells of Stage One or the subjacent cornified cells of the following stage will be encountered. Unpublished obser- vations by K. O. Haldeman show, in fact, that a cornification of the vaginal epithelium appears first in the solid epithelial core which extends from the vaginal lumen to the surface of the body. In this at separated intervals intra- epithelial vesicles are established. In cases where vaginal opening is coincident with cornified changes throughout the vagina and the animal is sacrificed twenty- four hours later, sections of the ovaries and oviducts will show a single set of corpora lutea, the first ones, and the first eggs in the Fallopian tube; and from what has been said of the long persistence of the corpora of ovulation we may be certain that no shortly preceding ovulation has occurred. Furthermore, animals killed before the opening of the vagina possess ovaries without any corpora whatever, but exhibiting merely follicles in various stages of growth and especially in stages of atresia. Ovulation therefore does not happen before the establishment of the vaginal orifice. These circumstances indicate that the first ovulation and the opening of the vagina often stand in very close relation to each other, but it would be a mis- take to take the latter as a criterion for the former. In fact, in animals in which the ovaries have been ablated at about the thirtieth day of life the vaginal orifice is established at about the usual time, so that this event need not stand in any relation to the first oestrus. In a considerable proportion of normal cases (as high as 40 per cent of all cases) the establishment of the vaginal orifice takes place before the first oestrus and hence by the breakdown of the closing mem- brane without cornification in the vagina; and for a few the samples of the vaginal content will disclose a smear consisting only of scattered epithelial cells with a few leucocytes. Sections of the ovaries in these cases before the incidence of oestrus show that an ovulation has not yet occurred; but when the growth, cornification, and dehiscence of epithelium takes place and is finally disclosed by the vaginal smear, follicular ripening and ovulation are invariably found. These first oestrous changes in the vaginal smear which are thus actually associated with the first ovulation, are not usually long delayed after the estab- lishment of the vaginal orifice, though we have encountered isolated instances of great delay. As will be seen from table 12, the majority of those animals 52 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA which do not experience their first ovulation simultaneously with the estab- lishment of the vaginal orifice, nevertheless do experience this ovulation within five days thereafter. Table 12. Showing chronological relation of first ovulation to opening of vagina in 193 rats. Instances Opening and first ovulation simultaneous 88 46% Ovulation within the first 5 days after opening 43 22% Ovulation 6 to 10 days after opening 22 11.5% Ovulation 11 to 15 days after opening 20 Ovulation 16 to 20 days after opening 4 Ovulation 21 to 25 days after opening 7 Ovulation 26 to 30 days after opening 5 Ovulation over 30 days after opening 4 Total 193 Thus, while our studies would lead us to caution any one who would attempt to use the first establishment of the vaginal introitus as an index of maturity, they nevertheless demonstrate that the first oestrus occurs within a few days succeeding this event, and that the first oestrus is not peculiar in the respect that any lack of correlation exists between ovary and vagina, for the correlation is identical with that which we have shown to obtain throughout the sexual life of the adult, i.e., oestrous indications in the vaginal smear are from the beginning associated with the incidence of ovulation. The second curve (B) shows the age at the time of establishment of the vag- inal orifice in 466 individuals. It will be seen that the orifice is established in most cases between the fifty-third and one hundred and second day of life, the average age computed from our data being 76.5 days. In a group of carefully selected, well-kept stock the average age was even slightly higher—about eighty days. Other groups have given an average much lower, so that we must admit that our data must be considered merely as depicting the conditions obtaining in our colony during the time occupied by this study. There are thus about ten or twelve days between the average age at first oestrus and the average age at first coitus, a fact which could perhaps find its readiest explanation in the coincidence of the latter date with the approximate time of occurrence of the second oestrus, for, as shown in table 13, the first oestrous cycle averages approximately ten days in length. In the majority of cases where the formation of the vaginal opening and the first oestrus are sim- ultaneous the closing membrane is not ruptured until Stage Three is reached, by which time oestrus proper has passed. This may explain failure to breed at the first oestrus in this group. LONG-EVANS: OESTROUS CYCLE IN THE RAT 53 OPENING OF VAGINA. NUMBER OF INSTANCES Opening of Vagina Average 76.5 days / . First copulation FIRST COPULATION. NUMBER OF INSTANCES Average 92.7 days DAYS AGE IN DAYS Curve showing distribution at five day intervals of ages at which first copulation occurred in 199 rats. Similar curve for ages at which the vaginal membrane ruptured in 466 rats. A surprising fluctuation exists in the age of maturity. While the extremes observed (table 13) indicate that rats may mature as early as at 45, or as late as at 147 days, there is also a considerable variation in the average age of maturity of a unit lot of rats handled at one time in exactly the same way as regards food, space, etc., and the same applies to litter mates. Some indication of the way in which the nutritive factor might tend to influence our results is suggested in the careful records of the time of maturity of 70 rats in which the usual table scrap diet was supplemented by whole milk and in some instances Table 13. Age in days at opening of vagina and at first ovulation, and length in days of first four cestrous cycles in a group of 200 rats. Range Average Age at opening 34th-109th day 72d day of life Age at first ovulation 45th-147th day 77th day of life Length of first cycle 3-37 days 10 days Length of second cycle 3-25 days 9 days Length of third cycle 4-26 days 8.5 days Length of fourth cycle 3-23 days 7.3 days 54 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA bv still other articles of diet, notably one-half gram uncooked liver daily. In these animals the growth rate was notably accelerated, but the age at maturity averaged 84 days, so that it appears that a superior nutrition need not hasten the establishment of ovarian function. VII. THE PHENOMENA OF REPRODUCTION AND THEIR EFFECT ON THE NORMAL RHYTHM A. Pregnancy 1. LENGTH OF GESTATION The exact length of gestation in the rat, which we would define as the time intervening between copulation and the occurrence of parturition, lies between 21.5 days and 22 days. In two hundred and one cases it was possible for us to know the period of gestation by quite accurate observations of the beginning and terminal event, that is, it was possible to know within eight hours when copulation took place by observations of the presence of the vaginal plug; and by placing animals in the obstetrical cage, which has been described earlier, the exact time of parturition was determined. In seventy-six of these cases (38 per cent) the time of gestation was within a feAv hours of 21.5 days; in one hun- dred and four cases (52 per cent) it was slightly less than 22 days. Table 14 shows the distribution of these instances of various gestation lengths. Some of these variations could be accounted for by variation in the relations between ovulation and copulation, as shown for embryos (p. 33). The longer ones must be attributed to other causes still unknown. In none of the above was the mother suckling a litter, a condition first shown by Daniel (1911) for mice, and King (1913) for rats, to influence greatly the span of gestation. Table 14. Length of period of gestation in the rat. (Probable error 8 hours) Length in days Instances 20 1 21 9 21H 76 (38%) 22 104 (52%) 22 2 23 7 24 2 Average of all gestations 21.8 days LONG-EYANS: OESTROUS CYCLE IN THE RAT 55 2. PROPORTION OF OVA PRODUCING YOUNG It is well known that resorption not infrequently occurs, even in fairly late stages of pregnancy. It is also conceivable that in many instances fertilization does not result or that some of the ova never become implanted. Data for the first time can now be presented giving an accurate idea of the per cent of eggs which do not produce young. From records of some 625 litters (table 15) it is found that the average nmnber in the litter is 6.9, and that the mode also is 7. As previously stated (table 4), the average number of eggs found in one oviduct is 4.8, an average of 9.6 for the individual at each ovulation. If all were fertilized and could develop, the average expected size of a litter should be 9.6. According to these figures, somewhat less than one-third of the eggs liberated from the ovary never develop. To be more certain, the average size of the litters was determined for the same group of animals in which those fifty individuals were taken in which the eggs could be counted in the oviduct. The average size in this group of one hundred and fifty-six litters (table 16) is 6.4, making a mortality of exactly one-third of eggs matured and liberated from the ovary. It would be interesting and important to determine the effect of nutrition on the number of young produced and whether the number of eggs matured could be increased, and also possibly whether the amount of atresia in the ovary less- ened. It is hoped to furnish such data at a later time. Such a problem should also be attacked from the standpoint of heredity. Table 15. Number of young per litterbin the rat. Number in litter Instances 1 1 2 8 3 27 4 58 5 70 6 110 7 114 8 105 9 68 10 43 11 14 12 12 13 3 14 1 625 litters— average 6.9 56 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Table 16. Number of young per litter in same group of animals which furnished material for counts of eggs and corpora per ovulation. Number in litter Instances 1 2 2 5 3 9 4 13 5 20 6 SO 7 28 8 21 9 18 10 7 11 2 12 1 156 litters— -average 6.4 The incidence of pregnancy has a peculiar interest to the student of the physiology of reproduction. Special changes occur in all the organs of repro- duction, changes which owe their origin to the complex series of events charac- terized by coition, fertilization, implantation, and other characteristics of the pregnant state. The cause of these may in some measure be analyzed as we continue to know more of the specific role played by each component of the reproductive system. 3. EFFECT OF PREGNANCY ON THE REPRODUCTIVE ORGANS Vaginal changes. It lias always been supposed that ovulation, as well as oestrus, is withheld in the case of all mammals during the period of gestation, but we are not aware of the existence of any rigid proof of this supposition. The demonstration that ovulation is always correlated with a precise set of changes in the histology of the mucous membranes of the genital tract, changes which can be detected in the living animal by means of the vaginal smear, gives us a chance to discover whether this “sign” of ovulation is ever encountered during the interval oc- cupied by gestation. In our experience no oestrous changes occur in the cell content of the vaginal smear throughout the period of pregnancy. The smear is always typified by the occurrence of scanty epithelial cells and leucocytes, to which, as will be explained later, may be added red blood cells during the first three days of the third week. After the extensive exfoliation of cornified cells, LONG-EVANS: OESTROUS CYCLE IN THE RAT 57 which begins with and immediately follows the copulatory act, no more cor- nified cells are encountered in the vaginal smear throughout gestation; nor do sections of the vagina taken at various times during pregnancy ever show a stratum corneum and the particular kind of typical, high, stratified, squamous epithelium which always goes with the possession of a eornified layer. To show the time of first appearance of film of red blood corpuscles on the floor of the vagina near the cervix. Table 17. Days after fruitful copulation 12 13 14 15 Total Number of instances in all of which the day of copulation was known 1 44 51 3 99 Number of instances in which the day of copu- lation was calculated from the day of litter- ing 1 28 46 13 88 Total number of instances 2 72 97 16 187 Per cent 1% 39% 52% 8% 100% On the other hand, the epithelium of the vagina during pregnancy comes to acquire a characteristic histology. This is also ultimately the case with the appearance to the naked eye of the vaginal mucosa, which usually about the fourteenth, and always by the sixteenth, day of gestation has a thick, velvety appearance quite unlike that seen in the non-pregnant condition. It is, indeed, at this time, also, that the vaginal speculum will disclose a bright red, bloody discoloration of the floor of the vagina, especially in its upper part near the cervix, a discoloration actually due to the presence of free blood upon the sur- face of the mucosa, as the careful removal of it by means of a swab, as well as the examination of it when removed by the spatula will show; so also sections (fig. 65, pi. IV) of the exudate in situ. It is extremely probable that we have to do here with a leakage of blood of uterine, and, presumably, placental origin, which has escaped through the cervical canal. In our opinion, this is the earliest infallible sign of pregnancy in the rat which may be detected in the living animal. It is highly interesting that once it appears it does not persist throughout the remainder of pregnancy, but is typically present for but three days (i.e., from the fourteenth to the seventeenth day), during which time it is hardly added to or renewed to any significant degree because the color changes from the bright red of fresh blood to a brown color through the well understood “ageing” of the blood pigment (table 17). This sign may rarely begin at the thirteenth day, 58 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA or, again, be present for the first time on the fifteenth day. It will remain for future work to detect just what changes in the pregnant uterus bring about this transitory hemorrhage. We should turn now to what we have described as a characteristic histology for the vaginal mucosa of pregnancy. We may state that this mucous membrane shows in the middle layer of its epithelial cells a characteristic vacuolarization so extensive as to characterize the entire vaginal canal in the last week of preg- nancy when the epithelium has also reached a considerable height. The change has not yet appeared on the eighth day (fig. 63, pi. IV), but is evident by the tenth (fig. 64, pi. IV), and reaches its greatest expression from the sixteenth (fig. 66. pi. IV) to the twentieth (fig. 67, pi. IV) day of gestation. It is detect- able first in the epithelium of the mucosal folds near the cervix and vaginal portion of the cervical canal. It is very remarkable that a squamous transfor- mation of the superficial cell layers does not ensue, but that the surface cells instead of being flattened are always fairly voluminous structures, columnar in form. Many mitoses are present and these account readily for the growth in the height of the epithelium, especially when we bear in mind that but slight loss occurs from the inconsiderable dehiscence. Along with the absence of cyclical changes, heat and copulation do not occur during pregnancy. However, our records show that two animals copulated during gestation, one on the fourth and fourteenth days, and the other on the sixteenth day. Ovarian changes. (1) Suspension of ovulation. Our demonstration that oestrous changes in the vaginal mucosa do not occur during pregnancy might, in view of the inva- riable correspondence which we have observed between the ovary and genital tract, be taken to indicate that ovulation is also withheld, but the final conclusive proof of the suspension of ovulation is afforded the employment of a selective vital dye for the corpora lutea. Some time ago one of us reported the elective behavior of certain vital benzidine dyes with respect to ovarian tissue, and called attention in particular to the fact that deposits of these dyes (adminis- tered to the living animal) were especially accumulated in the macrophages of atretic follicles and to a lesser extent in the lutein cells of most corpora lutea. We have experimented with eighty-three dyestuffs (most of them benzidine compounds) which by reason of their constitution might be expected to serve as vital dyes. Their behavior toward the ovary has been summarized in table 36 of the Appendix. We may remark here that most of these substances are competent to act as vital stains in the sense in which that term is usually em- ployed for the effects produced by Trypan Blue and other benzidine dyes. In the case of those dyes which produced a deep vital tinging of the animal as a whole a widespread vital staining was obtained—an effect on the cells of the LONG-EVANS: OESTROUS CYCLE IN THE RAT 59 connective tissue throughout the body and on certain of the endothelia—whereas in the case of those which do not tinge the body as a whole there is nevertheless always at the site of their injection an identical effect on the same cells, although one which is restricted to the immediate locality. Most of these eighty-three substances, which are thus ingested and stored by the macrophages of the living animal in granular form, are to he understood as segregated and concentrated in vacuoles which Evans and Scott have called “the segregation apparatus” of the cells, most of which may be specially formed for this purpose. The cells of the mesotheliiun, though differing from the endothelia, the fibroblasts, and the tissue macrophages, also participate in a characteristic way in this reaction. It is highly interesting to note that the most successful general vital stains, in the sense in which this word has usually been employed, fail to give the most satisfactory effects on the true lutein cells of the corpus luteum. Many of the dyes, such as Trypan Blue, or the first two dyes listed in table 36, are invariably brilliant general vital stains, and, while they tinge deeply the corpora lutea as a whole, microscopic investigation of the fresh tissue shows at once that the color is due largely to deposits in the mesothelial ovarian covering (the germi- native epithelium), and that the true lutein cells are conspicuous in contrast thereto by the possession of only very pale deposits of the dye. Indeed, Dye 26 of the table gives an intense crimson stain to the entire animal, hut produces granular deposits in the corpora lutea too inconspicuous or pale for detection! Nevertheless, it is undeniable that some general deep vital stains produce more easily distinguishable, deeper colored deposits in the lutein cells; and we would instance here dyes 9, 10, 14, 15, 41, and 59 of the table. We have discovered that two dyestuffs whose intensity of reaction with most vitally stainable cells is inferior to the above deep general dyes, nevertheless give the most distinctive and best colored deposits in the true lutein elements, and, above all, give the only deposits which, having these qualities, are also capable of perfect histological fixation. It is apparent, therefore, that the reaction to- ward these dyes on the part of the corpus luteum cells is somewhat different in character from that displayed by the great mass of those cells in the connective tissues and endothelia which are stained vitally, especially the so-called “macro- phage cells. ’ ’ These two vital stains, which, while giving merely a weak tinging of the animal as a whole, nevertheless so clearly mark with deposits the corpus luteum cells, are Africlol Blue, a combination of diclilorbenzidine diazotized and coupled in alkaline solution with two molecules of 1.8 amidonaphthol 3.6 disul- fonic acid, 60 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA AFRIDOL BLUE (DYE 11 OF TABLE 36) and Dianil Blue 2B, the combination of ortho-tolidine diazotized and coupled in alkaline solution with one molecule of 1.8 dioxynaphthaline 3.6 disulfonic acid (chromotrope acid) and one molecule of alpha-naphthol 4 monosulfonic acid (Neville-Winther acid). DIANIL BLUE 2R (DYE 51 OF TABLE 36) Afridol Blue produced by far the deepest colored deposits in tlie lutein cells, but after prolonged experimentation with it we decided that even in small doses its general toxic effect Avas so eAudent as to jeopardize one’s confidence that one is dealing with normal phenomena after its employment. In particular the reaction of this substance on the organs of the reproductive system justified a further search for a less toxic dye, for, administered during pregnancy, Afridol Blue Avas often responsible for abortion or resorption of the conceptus, and in other cases appeared equally responsible for an extensive follicular atresia so frequently encountered in the ovary after its administration. The second dye, Dianil Blue 2R, AA7as remarkably adapted to our purpose. This dye, in aqueous solution a deep reddish-blue, or bluish-violet, can be ad- ministered at any time in the sexual history of an animal in doses so large as to fill the peritoneal cavity and on four or five successive days Avithout serious toxic effect and usually, indeed, without any detectable toxic effect on the animal.3 It is but fair to state, however, that several points in the technique of the intraperitoneal ad- ministration of the dye are worthy of note if one would be sure to disturb least the normal physi- ology of the experimental animal. The animal should not be submitted to the toxicity of even transient anaesthesia and the dye administration should follow rather than precede the daily feeding time. It is somewhat surprising, but true, that animals which have been handled daily are sufficiently tolerant of manipulation to be grasped tightly in one hand by the skin of the back while the hypodermic needle punctures the ventral abdominal wall and the injection is completed. The distress occasioned by a voluminous intraperitoneal injection (e.g., 5 c.c.) is so great that rats will not partake of their daily ration if the injection is made within a short time of the feeding hour, but if the injection follows the feeding hour by a one to three hour interval the full stomach does not empty itself per os, and a normal nutrition is thus assured. s Reference has already been made to this vital ovarian dye in other work issued from this laboratory by Corner and Hurni and by Monroe Sutter, to whom we have introduced its employment. LONG-EVANS: OESTROUS CYCLE IN THE RAT 61 If, on the day following three or four daily intraperitoneal injections of this dye, an animal be sacrificed and the corpora lutea which happen to be present in the ovary examined in the fresh by cutting out a minute piece of the corpus with iridectomy scissors and crushing this in salt solution under a cover glass, it will be seen that each cell of the germinal epithelium contains considerable numbers of deep blue vital dye “granules” and that all the true lutein cells also contain smaller numbers of more scattered, minute, spherical vital dye “gran- ules.” The “dye granules” are of almost exactly the same dimensions as the lipoid spherules which occur in the cell at this time and with which they are Fig. 50. Outline drawing of fresh ovary of Rat 3514 (before fixing in Benda’s fluid) to show the corpora of pregnancy (deep blue), of lactation (light blue), and of ovulation (2 sets, colorless). Sections in figure 51, plate VIII, and figure 89, plate IX, are taken along lines AB and CD. Given dye during lactation; young weaned on the 25th day; the next Stage One occurred 3 days later; the following day in Stage Two a glass rod was introduced into the cervix; fifteen days later the animal was killed in Stage Three of the first recurring oestrus; eggs of this ovulation were found in the distal folds of the oviduct. admixed. The dye deposits, or granules, are fixed readily in formol, Bouin’s fluid, or Zenker’s fluid, but, unfortunately, are not visible in material preserved in mixtures containing osmic acid. In view of this fact, in all cases in which we desired to examine the lipoid character of corpora lutea which were at the same time marked vitally by means of these dye deposits, the ovary was sketched in salt solution under the camera lucida and the position of all corpora noted, the vitally stained corpora being especially labeled so that their identification in section was easy (figs. 50; 51, pi. VIII, and 89, pi. IX). In the ovaries of animals treated with such a vital dye only certain of the corpora are affected in the way in which we have just described; other corpora, which we have reason to believe 62 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA are younger, are unaffected. A careful test of this question in animals whose sexual and oestrous history was accurately known did, in fact, show us that cor- pora lutea, whether of ovulation, gestation, or lactation, do not readily receive deposits of the vital dye during the first three or four days of their life (figs. 59 and 60, pi. ArI). It is apparent, therefore, that the vital-stainability of the corpus cell, by which we mean its permeability to the vital dye substance and its segregation and storage of the same, is characteristic only of corpus cells of a certain age (fig. 60, pi. VI), the younger corpus cells (fig. 59, pi. VI) main- taining a completely refractory behavior toward penetration by the dye, just as was the case when they constituted the granulosa cells of the ripe follicle. Older lutein cells, then, constitute a different physico-chemical protoplasmic system and the vital dye gives a clear demonstration of how this change may occur in the cytomorphosis of a single cell. Corpora lutea which are produced subsequent to dye administration are always unstained. They never have traces of the vital dye, so that absorption of the coloring matter from an old corpus on the part of a fresh structure does not take place; nor have we evidence that vitally stained macrophages, which, as we know, are often present in the corpora and have considerable powers of migration, ever wander from a stained to an unstained corpus. The vital stain- ing method is hence of inestimable value in the case of an animal like the rat, where the most diverse conditions pertain with regard to the age and number of corpora lutea present at any one time. Since late in gestation the corpora can be deeply stained with the vital dye they can thus be distinguished with certainty from the corpora produced at subsequent ovulations which do not possess the dye, but which might resemble them in other ways. As will be shown farther on, this method is also essential for ready identification of the peculiar corpora lutea found in nursing animals, the corpora lutea of lactation, and has a more extended value in the fact that in cases where ovulation has been suspended after dye administration irrefut- able proof of this is seen in the fact that no unstained corpora arise in the ovary during the experimental interval. By this means we were able to demonstrate that the prolonged pause in oestrus experienced by rats after coitus or after the introduction of a rod into the cervical canal is accompanied by suspension of ovulation. If four or five cubic centimeters of a 1 per cent aqueous solution of this dyestuff be administered intraperitoneallv on the four successive days repre- sented by the fourteenth, fifteenth, sixteenth, and seventeenth days of gestation and the animal bred again at the incidence of the first oestrus (i.e., that which occurs on the day of parturition) and killed at the time of second parturition, one may obtain conclusive evidence of the absence of all ovulations during the second pregnancy. The ovaries of animals treated in this way contain only two LONG-EVANS: OESTROUS CYCLE IN THE RAT 63 kinds, or sets, of corpora lutea, one, tlie corpora of the first pregnancy, the only ones now present which are vitally stained and the only ones which were hence present at the time of dye dosage, and the other set, the corpora of the second pregnancy, normal in color, i.e., uninfluenced by the vital stain, and consequently to be interpreted as those which were formed subsequent to dye treatment. These corpora formed subsequent to dye treatment correspond in number with those which could be expected from a single ovulation (the average number of cor- pora lutea graviditatis found in one hundred cases of pregnancy being 4.9), and, moreover, in size and all histological characteristics, including the character of their lipoid deposit, they are identical and hence of the same age or ovulation, and constitute the corpora resulting from the ovulation on the day of parturi- tion transformed into the corpora of gestation on account of the second con- ception at this oestrus. The above facts appear to as to estalhish beyond peradventure the fact that ovulation is entirely suspended during pregnancy, vend, furthermore, demon- strate that the corpora of pregnancy result from the continued growth and enlargement of the particular corpora of ovulation from the follicles of which the ova were fertilized. (2) Size and characteristics of the corpora lutea graviditatis. Until almost the middle of pregnancy, i.e., during the first nine or ten days after a fruitful copulation, the corpora lutea do not differ in outspoken characteristics from the corpora of ovulation in their size or in the number, size, and distribution of the lipoid granules (fig. 68, pi. VII, 14th day) ; distinctive characteristics of the corpora of pregnancy are hence lacking. But after the tenth day of gesta- tion continued slow growth of these corpora permits them to attain dimensions never found in the corpora of ovulation, dimensions which may approach two millimeters in diameter and which toward the end of gestation (sixteenth to twentieth day) are almost always from 1.75 to 1.9 millimeters. This enlarge- ment is brought about to some extent, of course, by growth of the extra-luteal constituents of the corpus, but it is produced chiefly, if not entirely, by growth of the individual lutein cells comprising the structure. This final enlargement is not shared by corpora of the preceding ovulation; in other words, the latter are not capable of responding in this way, the corpora of gestation being pro- duced from the single set of follicles the eggs of which were fertilized. If copulation and pregnancy had not occurred the above set of follicles would have given rise to corpora of ovulation. This is easily proved in a case in which an animal, which was given dye during gestation, was allowed, after parturition, to pass one cycle, and was then bred at the second oestrus. The cor- pora of the first gestation, though small, were blue and could clearly be distinguished from the large corpora of the last gestation while all the others were of the one ovulation between. 64 MEMOIRS* OF THE UNIVERSITY OF CALIFORNIA Table 18. Comparison of different kinds of corpora lutea with regard to size of corpora and size and amount of lipoid granules Corpus luteum of Length of functional period in days Characteristics during first cycle Characteristics during end of first and be- ginning of second cycle Characteristics during third cycle Rat No. 3533 3533 454 Days 2 4 11pp. Ovulation 4-5 Diameter 1 mm. 1.05-1.2 1.1 Lipoid: Size* 1-2 1-7 (2x9) 1-7 Amount moderate moderately heavy Very moderate Fig. 48 49 Rat No. 3639 3587 3514 Days 10 13 15 (eggs in ovi- Copulation or 8-16 Diameter 1.35 1.1 duct) pseudo (about) Lipoid: 1.4 pregnancy Size* 1-3 1-5 (8) 1-9 Amount moderate moderate very heavy Fig. (placen- 88 toma) Rat No. 3479 3558 3471 447 428 452 Days 4 10 16 21 Within 12 hrs. pp. 7Yl pp. Diameter 1.15 1.22 1.8 1.7 1.75 1.23 Pregnancy 21^-22 Lipoid: ■ Size* 1-2 or 3 1-3 1-9 1-9 (2x9) 1-9 Amount moderate moderate moderate Very heavy Moderate Fig. 69, 70 71 Rat No. 366 373 425 3629 360 358 Days 9 14 21 2 after 4 after weaning 8 after weaning weaning (eggs in oviduct) Diameter 1.03 1.36 1.45 1.2 1.2 1.05 Lactation 21-45 Lipoid: Size* XA A 1-5 (8 & 9) 2-9 1-7, 9 Amount Very moderate moderate Very Very heavy Moderately moderate heavy heavy *The size is indicated by numbers which are given to the ocular spots shown in figure 46. See also p. 38. The extremes include the deeper half of the corpora. Figures in parenthesis are of relatively few, large, lipoid granules. Marked changes in the lipoid morphology of the lutein cells of the corpora of pregnancy, changes which we believe are indicative of cessation or diminu- tion of function, take place at the end of pregnancy; and within twelve hours post-partum (tigs. 69 and 70, pi. A7II), the enlargement of lipoid globules and hence great inequality in the size of these bodies as well as their increase in amount is very evident (see table 18). In the case of animals which are not permitted to suckle their young and which are not bred on the day of parturi- tion (i.e., cases in which pregnancy is succeeded by normal ovulation), these cells of the corpus of pregnancy continue to become tilled with lipoid deposits until the time of occurrence of the next ovulation following that on the day of LONG-EVANS: OESTROUS CYCLE IN THE RAT 65 parturition, when, just as we have shown to he the case with the corpora of ovulation, the lipoicl in the rapidly ageing lutein cells undergoes a second phase, a phase of marked diminution (tig. 71, pi. VII). (3) Persistence of the corpora lutea graviditatis. The method of staining vitally the corpora of pregnancy enabled us, of course, to follow the continued regression and final time of disappearance of these structures in animals killed at various time intervals after the gestation in which the corpora were stained. Although functional impairment of these bodies may already be detected at the close of gestation, they persist in the ovary as do the corpora of ovulation, but for a longer time. If the animal be sacrificed and the ovary examined four days after parturition, i.e., after the occurrence of the first post-partum ovula- tion (fig. 56, pi. VI), the corpora of pregnancy are seen to be stained vitally even more deeply than is the case in animals which are examined immediately after the conclusion of actual dye administration or at parturition (fig. 55, pi. VI). This apparent increase in vital stain of the corpora is, in our estima- tion, due to the fact that with the formation of subsequent corpora lutea, which protrude somewhat from the ovary, those of gestation slowly diminish and in this process their covering tissue, the germinative epithelium, to a certain extent follows them. Since the cells of the germinative epithelium are stained more deeply than the subjacent lutein cells, they make an even more emphatic color impression when now concentrated in the smaller area. This method of iden- tifying cells by means of the vital stain thus shows the retention of the identity of the mesothelial tissue overlying a particular corpus and demonstrates that most of the epithelial covering of the new corpora represents a new formation rather than a shift in position and stretching out of the mesothelial tissue which is vitally stained and once covered the old corpora. Twenty days after parturition the corpora lutea of gestation are still sub- stantial structures—somewhat over a millimeter in diameter—though the num- ber of their proper lutein elements is considerably diminished. Even about the fiftieth day after parturition (fig. 58, pi. VI) the corpora of gestation are about half a millimeter in diameter and contain numbers of lutein cells in which always a few small lipoid grannies are visible, though at this time by other methods the structure would hardly be recognizable as that of a preexisting corpus, for the inconspicuous lutein cells are separated by connective tissue in which vitally stained macrophages are prominent. By means of the vital dye method, however, it is possible to detect small areas in the ovary, usually seen from the surface as deep blue spots, which are the last vestiges of the corpora of gestation about one hundred and twenty-five days after the day of partu- rition. Though most of this small area is considerably beneath the surface of the ovary, still part of it usually reaches the surface; the macroscopic color, of course, is due solely to the vitally stained germinal epithelium and to macroph- 66 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA ages which have completely engulfed all lutein elements with their vital dye contents. The corpora of have thus been identified in the series of cases represented by figures 72, 83, plate VIII; 53, 89, plate IX; 54, 73, 82, plate X. Table 19. Length of first post-partum oestrous cycle when • suckling was prevented. Days Instances 3 2 4 2 5 11 97 6 4 7 7 8 1 9 0 10 1 11 . 0 12 0 13 3 14 2 27 cases average 5.5 33 cases average 6.9 days Table 20. Length of first post-partum oestrous cycle when young were nursed less than twelve hours. Days Instances 4 1] 5 4 6 3 *20 7 5 8 7. 15 1 *20 cases average 6.6 days. 21 cases average 7 days. (4) Effect of the corpora lutea gravidatis on other corpora lutea. We have already stated that on the twentieth day of the gestation period only a single conspicuous set of corpora lutea are found, the corpora of gestation, but that on the twentieth day post-partum, when normal ovulations have occurred, not less than thirty corpora are present in the ovary. It is apparent, therefore, that there is not only a suspension of ovulation during pregnancy hut also profound retrogression and resorption of all preexisting corpora lutea. All corpora of ovulation, then, regress, even those formed at the oestrous period immediately preceding that at which mating was permitted. LONG-EVANS: OESTROUS CYCLE IN THE RAT 67 The more substantial structures represented by corpora of a preceding gestation do not diminish to the same extent as do those of ovulation. In animals kept isolated, normal ovulation should succeed parturition at regular four or five day intervals, but we have found that the first oestrous cycle following parturition, as determined by the vaginal smear, tends to be somewhat longer. Table 19, compiled from 33 cases, shows that this cycle is 6.9 days in length. Table 21. Effect of number of young suckled on length of interval between weaning and first subsequent ovulation. Young weaned at 21 days. No. in litter No. of litters Average increase in weight of litter* Average length of interval 2 5 51 grams 3.4 days 4 7 59.4 grams 4 days 6 10 91 grams 6.9 days 7 8 100 grams 9 days 8 7 87 grams 7.7 days 9 7 88 grams 6.7 days 10 4 111 grams 10 days 11 2 106 grams 11 days 12 2 83.5 grams 10.5 days 13 2 94.5 grams 14.5 days *The litters were weighed at birth and at 14 days, and not at 21 days because during the last week the young are likely to partake of other food. We are indebted to Miss G. J. White for this data. B. Lactation 1. EFFECT OF LACTATION UPON THE VAGINA Young rats may usually be weaned after twenty-one days of suckling, but if left with the mother the duration of lactation may be double this time. Daily microscopic examination of the vaginal smear from suckling mothers shows the persistence of the condition characterizing the dioestrous interval, i.e., leuco- cytes and scattered epithelial cells. In the light of the invariable correspond- ence which we have found to obtain between ovulation and the character of the vaginal smear it might be inferred that the above evidence indicates that ovu- lation is suspended during suckling. This evidence is strengthened by the recurrence of the usual typical oestrous changes in the vaginal smear, which take place in about seven days after the suckling young are removed. The recurrence of the oestrous changes in the vaginal smear may take place from three to twelve days after the removal of the young and would appear in general to have some dependence on the number of young suckled. Oestrus recurs somewhat more tardily when the litter is large and the drain occasioned by lactation is consequently greater, an indication of which may be obtained from 68 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA the weight increase of the litter as a whole (table 21). When the young remain with the mother, oestrous changes in the vaginal smear will occur spontaneously at some time between the twenty-fifth and fortieth day after parturition. The histology of the vaginal mucosa during lactation shows that not only are the cornified changes associated with oestrus absent, but that a very consid- erable and characteristic reduction in the height of the epithelium occurs. While in the second day of lactation (fig. 74, pi. IY) this mucous membrane may still consist of four or five simple cell layers, by the fourth (fig. 75. pi. V) day more than three layers are seldom encountered, and on the sixteenth day (fig. 76, pi. Y), when lactation may be assumed to be at its height, most of the vaginal mucosa is actually reduced in its epithelial investiture to two cell layers. The superficial cells of the vaginal mucosa in lactation are cubical or low cylin- drical elements. The strict dependence of this characteristic epithelium upon the performance of the mammary glands could be illustrated in no more striking way than by the transformation of it which may take place within forty-eight hours (fig. 77, pi. Y) after the removal of the young when a high, stratified, squamous epithelium results. 2. EFFECT OF LACTATION UPON THE OVARY Suspension of ovulation. It is well understood that as a rule ovulation is suspended during lactation, an idea entirely verified by the findings in the vaginal smear which we have just detailed. However, proof that ovulation is inhibited during lactation may be furnished by staining the corpora of gestation vitally at some time during the last week of pregnancy and examining the animal so treated at any time during the first three weeks of lactation, when only a single unstained set of corpora lutea are found—the corpora of lactation. Furthermore, if in animals treated in this way we wait until the vaginal changes indicate the advent of the first oestrus after nursing, we find in these ovaries, besides the stained corpora of gestation and the unstained corpora of lactation, only the freshly formed corpora of this ovulation. It is, therefore, established that lactation inhibits ovulation. Corpora lutea lactationis. The corpora which we have designated the corpora lutea of lactation deserve distinction from the corpora of gestation and of ovulation on account of the char- acteristic picture of their lipoid bodies, for these uniformly distributed spherules or granules (fig. 79, pi. VII), which brown or blacken with osmic acid, are smaller than those possessed by the corpora of ovulation or of gestation. The corpora of lactation result from the post-partum ovulation occurring within twenty- LONG-EVANS: OESTROUS CYCLE IN THE RAT 69 four hours of parturition, as is also proved by marking the only other corpora which may be confused with them, the corpora of gestation, with the vital dye. Thus the lactating act, which has been in force at least twelve or fourteen days, imposes on the corpora of the post-partum ovulation a characteristic morphol- ogy of the lipoid deposits which, though slightly larger than the minute fat- bodies sometimes found in the granulosa cells before follicular rupture, are never permitted to attain the average size which is already reached by the lipoids of the ordinary corpora of ovulation (fig. 48, pi. VII) or gestation (fig. 68, pi. VII) within twenty-four to thirty hours after their formation. The corpora lutea of lactation, however, can not be easily distinguished in other ways from those of ovulation or gestation; they continue to grow beyond the size attained by the corpora of ovulation, although they do not attain the dimensions of the corpora Intea graviditatis; they apparently reach their maximum size at some time between the fourteenth and twentieth day of suckling, when a diameter of 1.4 millimeters may be attained. We have already mentioned the fact that, to judge by changes in the lipoids, the impairment of the corpora lutea of ovulation barely precedes the next oestrus, but so short is the time interval involved between the beginning of these changes in the ordinary corpora (ovulation) and the next oestrus that one might have difficulty in assuring himself as to which of the two events is anterior. We have also shown that this difficulty is not experienced in the case of the re- gression of the corpora of gestation, for well marked changes in the lipoids of these cells are always elegantly displayed in the first twelve hours after the birth of a litter and before the post-partum ovulation has occurred. The relation of lutein cell degeneration to the incidence of the next oestrus is even clearer in the case of the corpora of lactation. Cessation of lactation provokes lipoid changes within the lactation corpora within twenty-four hours so that we have in our hands the means of bringing this about simply by re- moval of the young. One day after the litter has been removed the lactation lutein cells no longer possess their fine, dustlike lipoid granules (fig.. 79, pi. VII), but have lipoid granules which are already larger than those found in the healthy corpora of ovulation or those of pregnancy. (See also table 18.) Two days after litters have been removed the lactation lutein cells contain still larger and markedly irregular lipoid masses, about like those at four days (fig. 80, pi. VII), and yet the next oestrus does not usually occur until at least four days reckoned from the removal of the young, by which time the lutein cells are gorged with lipoid granules crowded together and beginning to fuse into large masses, especially in most of the cells in the inner half of the corpus. One fact is evident from the sequence of phenomena which we have just described, namely, that degeneration of functional corpora lutea does not itself bring on the next oestrus, even though there can be no doubt, of course, that the 70 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA occurrence of oestrus is effectually restrained when corpora are functioning actively and that this restraining influence is released on the advent of decay of the corpus. It is evident that the fundamental cause of oestrus, whatever it may be, can not again operate so quickly after a period of lactation, and especially the lactation of a large litter, as it can under the normal conditions characterized merely by the usual succession of ovulations. It would be im- portant to assemble quantitative information here on the influence of lactation on the repression of the growth of follicles or on the production of follicular atresia. It is somewhat surprising that the termination of pregnancy, where, moreover, we have to do with a great bulk of active lutein tissue, permits a speedier recurrence of oestrus and ovulation, as witnessed by the prompt post- partum ovulation, and we can only explain this by the assumption that lutein cell decay has occurred anterior to parturition. Effect of lactation on preceding corpora lutea graviditatis. We have already called attention to the fact that on the last days of gestation the corpora lutea of pregnancy are the only corpora present in the ovary, the corpora of all preceding ovulations having been brought to atrophy and re- sorption. A somewhat similar effect is seen in lactation, during which the corpora lutea of the preceding pregnancy tend to disappear more rapidly than under conditions in which the young are removed as soon as born and pregnancy is succeeded merely by normal ovulation. When animals are placed in the obstetrical cage during the last day of pregnancy so that lactation is prevented, the young being separated from the mother automatically as soon as born, and when after these conditions the ovary is examined from the twentieth to twenty-third day post-partum, normal ovulations having intervened, the corpora of pregnancy are still seen as fairly substantial structures barely over a millimeter in diameter with many, though atrophied, lutein cells in which are moderate lipoid deposits (Rat 325). When the corpora of pregnancy are examined at a similar time interval, which has, however, been occupied by vigorous lactation, they are somewhat smaller in size and contain either no lipoid or very fine lipoid deposits in their lutein cells, which are still scantier in number and, indeed, in some cases have disappeared (Rat 425). C. Copulation 1. NORMAL COPULATION AND THE FORMATION OF THE VAGINAL PLUG From the preceding sections it will be seen that the length of the ovarian cycle may be influenced by various conditions. It is interesting that the act of copulation itself also exerts a similar though not so great an influence. An attempt to analyze this phenomenon has led to the discovery of a still more remarkable set of facts which can best be presented after a discussion of normal copulation. The manner of copulation in rodents and especially in the rat has been de- scribed by earlier investigators: Lataste, Steinach, Kirkham. Yet it may not be LONG-EVANS: OESTROUS CYCLE IN THE RAT 71 amiss to present a more complete account of certain aspects of the subject. The condition of the female is of importance because, as has been pointed out earlier in the paper, copulation can take place only when the female is in heat, which, as we have shown, occurs typically in the transition from Stage Oneto Stage Two of thu vaginal smear andm the early part of Stage Two. It will he recalled that at this time the lips of the external orifice of the vagina are slightly swollen and tend to make the opening more prominent; the vaginal mucosa is dryish, and covered with the thick cornified layer which, as Lataste long ago pointed out, probably serves as a protection. The cornification is accompanied by and possibly is the cause of a somewhat disagreeable odor which not unlikely is a means whereby the male is made aware of the condition of the female, for, as is common among animals, recognition may be chiefly carried on through the olfactory sense. The condition of the vaginal smear is sufficient to enable the investigator to determine in nearly all cases when the female will copulate and is of the greatest service in breeding animals for purposes of embryological study, for it is only necessary to look through a colony to pick out those indi- viduals which will usually copulate with little delay. In investigating the re- lation between the stages of the oestrous cycle and heat it was, of course, neces- sary to resort to the use of males, for many individuals show no signs that enable one to recognize the heat condition. Careful observation and consider- able experience has enabled us, however, to detect oestrous quickly in most cases by the behavior of the female when placed with males. It might be mentioned parenthetically that the female rat, unlike other mammals, only rarely attempts to play the part of the male in riding other females. When such a female in heat is placed in a large cage in which males are kept she does not remain quietly in a corner, but is more or less active, constantly moving about and often keeping near the males. If she is not in heat she does not respond in any way to the male and after a short time is ignored. On the other hand, if in heat, her reactions are characteristic and consist in running about more or less intermittently with a curious hopping gait, stopping when a male succeeds in making an attempt, when the back is flattened in a characteristic way; indeed, the back is bent so that it becomes concave with the tail up and the head back. Under these cir- cumstances and also frequently when she is hopping about, the head is shaken so that the ears quiver with a fine vibrating movement. It is an interesting fact that the position of slight opisthotonos taken by the female at coition may often be elicited by inserting rather firmly a speculum into the vagina while the animal is held in the hand, but the elicitation of this reaction has not resulted on account of contact of the speculum with the actual cervical canal, and unlike cervical stimulation it does not delay the appearance of the next oestrous cycle. This can not- be considered a sign of heat, for in some individuals the animal will respond in this way at any time during the cycle. Nevertheless, it is usually characteristic only of the period of oestrus. 72 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA One of the most striking features of the copulatory act in rats is the great celerity with which attempts are made and actual insemination is accomplished by males, the movements of the posterior part of the body being very rapid. After each attempt, of which there may be very many before a fruitful coitus is consummated, the male almost invariably rolls back on to his hind quarters and licks his genitals. To an inexperienced person these attempts may be easily mistaken for a true copulation. But experience and vigilant observation will show that the latter differs from the former in being slightly more prolonged and in the male more often rising on his hind feet instead of promptly rolling backwards. He is usually quiet for a time afterwards, but as such inactivity often follows abortive coitus it is not a reliable sign. The only certain indica- tion is the presence of a “plug” in the vagina of the female immediately after. Unless this is looked for at once it may be lost if the female is left with males for the reason that males may copulate as many as five times within a short period of time, and with the same female, one or two attempts being sufficient to dislodge a plug from the vagina. If the female is isolated the plug will usually remain in the vagina about twelve hours, occasionally as long as twenty-four hours. Of course, in any case, the actual finding of sperm in the vagina is the crucial test of insemination. But in our experience a plug is never unaccom- panied by sperm. The vaginal ping of rodents has been the subject of study by a number of investigators with regard to its structure, origin, use, and chemical composition. Although its occurrence was known before the time of Lataste and has been mentioned by many others since, he seems to have been the first to give it careful attention in his papers from 1882 to 3893. Lataste called the ping the houchon vaginale, and wras aware that most of it came from the seminal vesicle. He believed, however, that the outer part, or “envelope,” was a contribution from the vagina of the substance which we now know to be the cornified layer which in some instances was so abundant as to give rise to a large mass, in consistency not unlike a plug. Lataste’s conception of an outer “envelope” or coat formed by the vagina is easily explained by the fact that most plugs which are allowed to remain in position until they fall out do carry with them various portions of the cornified layer of the vagina, to which the plug is adhering tightly, the cornified layers being in the act of dehiscence at this time. Indeed, since one may occasionally even withdraw a more or less perfect cast of the vagina in the form of a sheet of cornified cells which constitutes the entire stratum corneum removed en masse, it is easy to believe that many of the vaginal plugs which Lataste examined must have contained an external coating with these cornified cells. Walker (1910) has shown conclusively that the plug is derived from the secretion of the seminal vesicle coagulated by the secretion of a coagulating gland (as the writers can confirm). As sperm are being expelled the secretions LONG-EVANS: OESTROUS CYCLE IN THE RAT 73 of the seminal vesicles and of the coagulating gland mix and evidently “set” the ejaculated mass, making it adhere tightly to any object with which it is in contact. One will have no difficulty in seeing the formation of a plug that may be withdrawn from the penis of an animal killed by a blow on the head. On the tip of such a plug is a small mass of spermatozoa, of which use has been made in artificial insemination. It might also be noted that occasionally a plug is found attached to the hair of the female. All these plugs have the same general shape. It is apparent that at the moment of ejaculation the plug is moulded to fit the vaginal lumen, especially at the cervix, for the plug is constant in shape Fig. 84. Diagram of longitudinal section of vagina with plug in situ extending a short distance into the cervical canal. Eat 3580. X 6. and has two little prolongations, each of which extends into one of the cervical canals (fig. 84). The plug adheres so closely to the cornified layer of the vag- inal mucosa (fig. 85, pi. XI) that it is not easily distinguishable in section. The plug is naturally loosened so that it may be dislodged from the vagina when this cornified layer becomes stripped off spontaneously in every oestrous cycle, as described in an earlier section. In fact, the cornified layer may not only be considered a protection but also a means of insuring loss of the plug after it has performed its function. Sometimes two or even three plugs remain in the vagina at the same time. The plug is not a solid body, but contains cavities, especially at its deeper end, these cavities being filled with spermatozoa. One of the obvious functions of the plug apparently is to confine spermatozoa between itself and the uterus as though to give abundant opportunity for them to make their way into the uterus. 74 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Immediately or very soon after copulation the uterus is found to contain a considerable amount of dark reddish fluid in which there are enormous numbers of spermatozoa. The presence of this fluid has been noted by other investigators (Sobotta for the mouse and Konigstein for the rat), but its relation to the fluid produced during Stage One has never before been understood. It will be re- called that in the first stage of the oestrous changes the uterus becomes dis- tended with fluid. But this fluid, unlike that after copulation, is always, in normal animals, clear and colorless. That the bloody nature after copulation is not due to the spermatozoa is proved by the fact that after copulation by vasectomized males, which are unable to eject spermatozoa, the same bloody appearance is to be noted. It is probably caused by the act of copulation itself. The male is capable of copulating and forming these plugs at any time of the day or night, but is most active at night. This last peculiarity may be the reason for a certain amount of apparent infertility in the case of females which are in heat for the first time in the morning hours and are past the heat period by night, when sexual activity in the male is at its height. In the breeding of rats for embryos for any purpose except highly accurate studies on age, the observation of the plug is of the greatest value in saving time. It is only necessary to examine daily with the aid of a speculum the vaginae of mated females early in the morning or late at night for plugs from which to date the approximate age of embryos. Many animals can in this way be examined quickly. From the variations found in the development of embryos, even of the same copulation age, a fact Avhich has been pointed out in earlier parts of this paper, it will be evident that some variation is unavoidable. But the inexactness in estimating the development of embryos dated from the finding of a plug is but slightly greater than if copulations be actually observed. Some plugs will prove to be infertile, of course, just as is the case with observed copu- lations. The following figures are of interest in this connection. One hundred and ninety-eight healthy young females were confined in several lots in each case with an almost equal number of healthy young males, their litter mates. During a fixed time interval we made daily observations with the vaginal speculum at an early hour each morning. Two hundred and forty-four gesta- tions resulted during this time interval and three weeks thereafter, but during the same time interval only one hundred and fifty-two plugs had remained in the vagina long enough to he ob- served in the morning round of examinations, and of these one hundred and fifty-two plugs, one hundred and thirty-four, or slightly over 88 per cent, were followed by conception. It is evident that by this method alone, regardless of the fact that many plugs fall out during the night, a copu- lation record may be obtained for many pregnancies, and in almost 90 per cent of all cases where the plug has remained in situ until morning, gestation will follow. On the other hand, it is to be pointed out that in a colony handled in this way we will have almost as many pregnancies which are not preceded by the observation of a vaginal plug in situ the following morning. In our par- ticular set of animals, of the two hundred and forty-four gestations resulting, in only one hundred and thirty-four cases was the plug seen the next morning, i.e., in only 54.8 per cent of the total gestations was it possible to obtain a record of the copulation by this method of observation. This figure might be higher if twice daily examinations were made.4 4 This is a far more expeditious method of dating embryos than by actually observing copulation, and might well be an important factor in deciding whether to use rats or pigs for laboratory work in embryology. LONG-EVANS: OESTROUS CYCLE IN THE RAT 75 2. INFERTILE COPULATIONS Diseased females with normal males. It has just been pointed out that a small percentage of copulations (or plugs) are sterile. It had been found early in this investigation that some of such females when left with males would copulate only at fairly long intervals with- out ever becoming pregnant. Such females turned out to be sterile because of an infection of the oviduct which occluded the lumen, but did not prevent ovu- lation in most cases. The reason for these long cycles is undoubtedly the same as for those which will be next reported. Table 22. Infertile females mated with normal males with intervals between plugs. Designation of Animals Intervals in days between plugs in order of occurrence 158 7, 3, 3, 3, 3, 5, 6 159 13, 26, 14, 13, 13, 12, 12, 13 160 17, 13, 16, 15 161 14, 13, 13 197 12 198 10, 13 201 20, 13, 4 202 27 205 17, 23, 19, 16 206 15, 25, 13, 27, 12, 13, 23 208 13, 25, 13, 12, 14 209 4, 2, 17, 16, 18, 16, 18, 17, 40, 19 211 14, 10, 7 212 3, 3, 3, 4, 5 214 17, 16, 14, 15, 15, 15, 14, 14, 12, 4, 14 247 18, 17, 17 249 28, 13 294 17, 19 310 13 316 13, 13 339 5, 12, 4, 6, 7, 2, 2, 3, 3 342 7, 8, 4, 3 361 9, 15, 11, 4, 4, 3, 2, 4, 3, 4 386 6, 10, 8, 3, 5 393 9, 3, 7 394 23, 5, 10, 12, 13 395 9 396 14, 13, 14 404 18, 16 406 14, 12 407 12, 13 431 10 441 14 450 12, 12, 12 1188 12, 26, 13, 10 76 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Normal females with vasectomized males. In the course of a study of living eggs by one of us (Long, 1912), it was found that when females were allowed to mate with males rendered sterile by vasectomy, eggs could always be procured in the oviducts during the day on the morning of which a plug was found. Such males, from which a small piece of each vas deferens was resected, could copulate in a perfectly normal manner, but were impotent to inseminate a female. It seemed then as though the use of these males furnished an excellent method of studying the occurrence of ovulation since if copulation took place only at the time of ovulation the pres- ence of the plug, uncomplicated by pregnancy, would he taken as an indication of ovulation. Females were accordingly placed each day with a new vasectom- ized male to give the maximum opportunity for coitus and careful daily obser- vations made. The results of many of these cases are shown in table 23, from Nine normal females mated with vasectomized males, with intervals between plugs. Table 23. Series Number Intervals between plugs’ (in days) in chronological order 142 12; 24, 11 301 13, 13 305 19, 14, 14, 16 309 7, 4, 13, 13 410 12, 28 413 12, 16, 27, 15, 26, 16 422 14, 4, 15, 16, 16 423 8, 4, 4, 4, 12, 16, 16 440 10, 4, 4, 18, 15, 14, 14 which we would have to deduce an ovulation period that could not be harmon- ized with the earlier supposed cycle of ten days, nor with the true later discov- ered cycle of four to five days. This long copulation interval of healthy females with vasectomized males was in complete accord with the results on the copu- lation of sterile females possessing healthy ovaries but occluded oviducts. Furthermore, in both kinds of cases the vaginal smear showed a delay in the appearance of the next oestrus (table 24). It was then that the suggestion offered itself that perhaps the secretions of the male accessory glands might be of importance in this connection. Previous studies had been made by other authors on the influence of prostate secretion on the activity of the spermatozoa and on fertility (Hirokawa, Iwanoff, Stein- ach, Walker), none on the influence on the female. Since it seemed possible that the plug itself, perhaps by being partly dissolved and absorbed, might exert an influence independent of the act of copulation, an attempt was made to mix 3. MECHANICAL STIMULATION OF CERVICAL CANAL LONGr-EVANS: OESTROUS CYCLE IN THE RAT 77 Table 24. Length of cestrous cycles following infertile copula- tions, or copulations with vasectomized males, as determined by vaginal smears. Length of cycle in days Number of cases 3 i 5 2 6 1 9 3 10 5 1 11 14 12 8 13 5 51 cases 14 4 average 15 5 13.1 16 3 days 17 2 18 2 19 3 - 58 cases—average 12.4 days the two secretions responsible for its formation in the vagina. Although a sort of plug was formed, but not successfully, no lengthening of the cycle followed. The only other possibility that suggested itself was the secretion of the pros- tate (table 25). Accordingly, the secretion obtained by triturating the bull’s prostate gland, and also extracts of it in Ringer’s solution, were injected into the uterus through the cervical canal by means of a glass syringe, using the method employed by one of us in earlier work (Long and Mark, 1911) wuth the result that the following cycle was longer than normal by several days, a very gratifying sequel. Controls consisting of injection of the same substance into the peritoneal cavity were negative. The same result followed, however, when Ringer’s solution was injected into the uterus and in a few cases when the syringe was inserted but obstructed in some way so that the contents could not be ex- pelled into the uterus. In fact, it was then discovered that merely inserting into the uterus a small glass rod had the same result! (table 26). In all of these cases the uterus was reached, of course, by traversing the cervical canal. As further experiments proved, the same results follow when the rod is inserted only about one millimeter into the cervical canal (table 26). This was done by using a slender brass rod with a small collar near the tip to prevent the rod being pushed in too far. It was apparent, therefore, that not only was it unnecessary to traverse the entire cervical canal but that the characteristic effect of cervical simulation could be elicited from that part of the canal known to possess the same stratified squamous epithelium as the vagina. The prolongation of the cycle follows only when the instrument is introduced during Stages One to Three 78 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA (table 26)., that is, during, and shortly before and after, the period of heat. This undoubtedly explains the action of the plug in prolonging the cycle since the plug fits tightly into the vagina and extends a short distance into the cervix. This discovery of the profound influence on the length of the oestrous cycle of stimulating the tip of the cervix of the uterus by the mere momentary intro- duction into it of a slender rod is one of the most remarkable reactions known to us. The questions which naturally arise are what is the explanation of the mechanism of this striking phenomenon, and what is its significance? A par- tial answer to these questions can be given when dealing with the subjects of deciduoma formation and ovarian transplantation. Table 25. Effects of administering various substances in utero or intraperitoneally. Figures in italics indicate the length of the period immediately following treatment, the other figures the length of periods preceding and following this. Secretion of rat prostate. In utero. Designation of animal 2412 4, 4, 15, 3, 4 Extract of rat prostate in Ringer's solution. In utero. About .2cc per dose. 2010 5, 4, 17, 5, 8 2332 4, 4, 15, 4, 6 2395 6, 5, 18, 4, 4 2518 3,20, 5, 3 Extract of rat prostate in Ringer’s solution. Intraperitoneally. 2383 5, 6, 5, 6, 6 Given in middle of period, lcc. 2395 4, 4, 6, 5, 11 Given on second day of period. .2cc. 2412 4, 4, h 3, 2 Given on second day of period. .6cc. 2557 5, 5, 6 Given on second day of period. l.Occ. 2547 5, 5, 19, 8, 3 .2cc. 2568 4, 5, 4 .2cc. 2897 6, 4, 5 .2cc. 3096 5, 6, 7, 7, 7 .2cc. .9% salt solution intraperitoneally. Doses of 4cc. 2208 5, 5, 8, 3, 6 2383 5, 4, 5, 6, 6 2502 5, 5, 5 2537 5, 5, 10, 11, 8 2571 9, 4, 4, 4, 5 LONG-EVANS: OESTROUS CYCLE IN THE RAT 79 Extract of seminal vesicle of bull in Ringer’s solution. In utero. ca .2cc 2319 8, 7, 21, 12, 5 Extract heated minutes at 60-62° C. 2518 3, 5, 5, 4, 6 Given on second day of period (about 24 hours after Stage 2). 2586 4, 5, 11, 5, 6 2850 6, 5, 13, 5, 5 Extract heated minutes at 60-62° C. 2861 5, 10, 5, 4 2896 8, 6, 6, 6, 7 Extract heated minutes at 60-62° C. 2908 4, 4, 7, 5, 8 Extract heated minutes at 60-62° C. Given late in Stage 2. 2930 5, 12, 4, 5 2975 5, 5, 11, 4, 4 3017 5, 18,12, 4, 4 Given on second day of period (about 24 hours after Stage 2). Extract of prostate of bull in Ringer’s solution. In utero. 2384 5, 5, 12, 4, 4 2502 5, 6, 10, 4, 5 2568 4, 5, 16, 5, 4 2917 3, 6, 12, 4, 4 3011 6, 4, 11, 3, 5 3095 6, 4, 10, 4, 6 Ringer’s solution in utero. 2319 12, 5, 21 2332 5, 5, 17 2350 5, 6,12, 3, 9 Failed to inject solution, but introduced syringe. 2351 8, 4,12, 4, 4 Failed to inject solution, but introduced syringe. 2384 4, 3, 13, 4, 4 2486 5, 5, 16, 3, 5, 4, 3, 17, 4 2518 4, 5, 17 2542 5, 6, 11, 4, 3, 6, 6, 12, 4, 6 2547 5, 5, 6 Tried to inject solution but failed. 2560 6, 5, 11, 6 2568 4, 4, 19, 5 2586 9, 5, 14, 4 2739 4, 4, 13 2850 5, 4, 14, 4, 4 2861 8, 4, 18, 4 2896 4, 4, 14, 5 2908 5, 4, 14, 5 2912 5, 6 2930 4, 8, 21 2942 5, 12, 5, 4, 6, 6, 11, 4, 4 2956 5, 4, 12, 6 2960 4, 5, 11, 5, 4 2963 6, 5, 16, 6 2968 5, 5, 14, 4 2975 21 2981 8, 7, 15 3011 4, 4, 21 3014 4, 4, 3, 14, 4, 4, 5, 15, 5 3017 4, 5, 12, 4, 4 3095 4, 4, 72, 7, 5 80 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Table 26. Length of oestrous cycle in days after introducing a glass rod once into the cervix a distance of about 10 mm. and allowing it to remain for a few seconds, during Stages 1, 2 or 3. Figures in italics indicate the length of the period immediately following treatment; the other figures the length of periods preceding and following this. Designation of animal Periods Remarks Designation of animal Periods Remarks 2350 5, 15, 4 3753 13, 12, 5 2384 4, 16, 4 4, 5,4 2486 4, 16, 4 4, 4, 4, 5 2518 4, 16, 5 5, 4> 4, 3 2542 11, 5 4, 4,4 2850 4, 18, 4 4, 4, 16, 5 2861 4, 15, 4 4, 4,4 2896 5, 12, 5 3754 11, 9, 5 2908 5, 12, 5 5,5,7 2942 4, 18, 4 7, 5,5 2956 6, 12, 4 4, 6,4 2960 4, 19, 4 4, 14, 8 2963 6, 11, 4 3755 8, 4, 5, 6 2968 4, 20, 3 5,14 Failed to introduce rod 3011 4, 16, 5 4, 4,4 3014 5, 16, 4 3756 7, 5, 4 Failed to introduce rod 3017 4, 16, 3 4, 5, 4 Failed to introduce rod 3028 7, 18, 5 4, 13, 3 3073 6, 19, 9 3820 16, 4, 4 3079 8, 15, 5 4, 18 Rod inserted only 1 mm. 3086 5, 18, 6 20, 12 3579 4, U, 4 3883 5, 13, 19, 11, 10 4, 5,4 Rod inserted only 1 mm. 6, 17, 6 4, 4 Rod inserted only 1 mm. 6, 15, 4 16, 4 Rod inserted only 2 mm. 4, 4, 14, 5 3599 4, 19, 4 4, 4,8 4, 21, 5 Rod inserted only 1 mm. 3886 5, 4, 4 3641 11 4, 4,4 3642 11 3887 7, 13, 13, 4 3723 4, 11, 5 Rod inserted only 3 mm. 4, 6, 7 3724 4, 13, 4 3888 5, 6, 4 6, 5, 11, 7 Rod inserted only 1 mm. 6, 20, 5 3726 4, 12, 5 Rod inserted only 3 mm. 3,4 3729 17 3890 6, 4, 12, 4 3730 13 4, 4,4 3746 7, 18, 4 3893 4, 4, 9, 16 4, 10 7, 18, 5 4, 12, 4 5, 15, 5 4, 11, 18, 4, 10, 4 4, 11, 4 4 5, 17, 5 3747 7, 10, 9 Rod inserted only 3 mm. 3895 6, 10, 10 9, 10, 6 3896 6, 3 6, 10, 4 3, 4, 12, 16, 5 6, 12, 5 4, 18, 4 5, 14, 4 4, 20, 9 4, 12, 11, 4 5, 18, 5 4, 13, 4 3898 4, 17, 4 LONG-EVANS: OESTROUS CYCLE IN THE RAT 81 Table 26—(Concluded) Designation of animal Periods Remarks Designation of animal Periods Remarks 3748 4, 5,3 Rod inserted only 3 mm. * 3898 4, 23, 16,4 3, 12, 5 4, 22, 4 4, 15 4, 19, 5 4, 13, 5 3899 5, 14, 4, 4, 4 3749 9, 4,10 5, 5, 10, 4 4, 11, 4 4, 19, 5 3750 5, 8, 6, 4 5, 6, 4, 6 9, 3 3900 4, 3, 10 4, 16, 4 3, 20, 4 4, 5, 8 4, 13, 4 8, 3, 13, 3 4, 14, 4 4, 13, 4 4, 13, 4 5, 4,4 3, 16, 4 3751 3, 15, 4 Failed to introduce rod 3901 4, 4,5 4, 11, 4 Failed to introduce rod 4, 4, 14, 3 4, 14, 4 4, 4, 6, 23, 4 4,5 4, 4, 20, 4 4, 15, 5 4869 8, 8, 6 5, 12, 5 4870 4, 4, 25, 4 4, 15, 4 4, 15, 4 Rod inserted only 1 mm. 4, 14, 5 Rod inserted only 1 mm. 4872 4, 15, 4 4, 16, 4 Rod inserted only 1 mm. 4, 13, 4 Rod inserted only 1 mm. 4873 5, 20, 6 Rod inserted only 1 mm. 6,17 Rod inserted only 1 mm. 4874 4, 20, 5 4, 4, 18 Rod inserted only 1 mm. 4875 4, 13, 5 4, 17, 4 Rod inserted only 1 mm. 4, 8, 5 Rod inserted only 1 mm. 2856 5, 14, 4 Rod introduced during interval 2739 4, 4, 15 Rod introduced during interval 2896 5, 4, 5 Rod introduced during interval 2908 5, 12, 4 Rod ntroduced during interval 2942 4, 4 Rod introduced during interval 2956 5, 6, 4 Rod introduced during Stage 3 2963 4, 14 Rod introduced during interval 2975 4, 5, 21 Rod introduced during interval 3014 4, 12, 4 Rod introduced during interval 3046 6, 15, 5 Rod introduced during interval 3054 4, 17 Rod introduced during interval 3064 5,11 Rod introduced during interval 3072 6, 4, 13 Rod introduced during interval 3082 7, 6, 14 Rod introduced during interval - Rats 3746 to 3756 had mammary glands excised (see table 31). Rats 3579, 3599, 3820, and 4869 had uteri removed (see table 25). Rats 3883 to 3901 were used for ovarian transplantations (see table 27). 82 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA 4. PROOF THAT COPULATION AND CERVICAL STIMULATION INDUCE A CONDITION OF ‘ ‘ PSEUDO-PREGNANCY ’ ’ At first sight the remarkable effect of copulation or of cervical stimulation in delaying the appearance of the next oestrus and ovulation would appear to be unexplainable, for these procedures would appear thus only to express them- selves some days after their occurrence and such a belated physiological re- sponse, or effect, would be difficult to understand. It would be easier to under- stand the delay of an immediately impending ovulation produced by nervous or other influences or the acceleration of that ovulation by direct nervous or vaso- motor means, but, as wre have shown, copulation or cervical stimulation seems not to influence the progression of oestrous changes and ovulation at the time, but only the time of appearance of the next oestrous events. The entire matter be- comes more intelligible, however, tvhen we discover that the effect of copulation and cervical stimulation upon the reproductive system is immediate, although its manifestation is deferred. The stimulus produces a condition which justifies the designation “pseudo-pregnancy.” Effect upon the vagina. During the ten or twelve day interval which elapses after infertile coitus or after stimulation of the cervical mucosa, the vaginal smear indicates that we have to do with the condition of prolonged dioestrum. During this time merely leucocytes and scattered epithelial cells are present in the vaginal smear. Furthermore, sections of the vagina confirm the inference which the smear would give regarding the absence of cornification changes in the mucosal epi- thelium. The vaginal mucosa does not increase in the characteristic way in height and stratification as it does immediately preceding oestrus, but, on the other hand, begins slowly to undergo changes which we have already described as characteristic for the condition of pregnancy, the main features of which are the occurrence of cuboidal or cylindrical surface cells and of vacuolization in the intermediate cell layers. Our material indicates that these changes are not apt to be so pronounced in animals killed at any time interval after copulation or cervical stimulation as they are at a corresponding time in pregnancy; nor have they, in our experience, progressed much beyond what is found on the tenth day of pregnancy at the most. The last statement may be amplified by the statement that the entire vaginal mucosa after cervical stimulation does not undergo the “pregnancy” changes, but that only the upper portions of it near the cervix do so. Figure 86, plate Y, shows the condition of the vaginal mucosa in folds near the cervix on the thirteenth day after insertion of a small glass rod into the cervical canal. The histological picture is not quite so far advanced as that of the thirteenth day of pregnancy, but, on the other hand, would seem LONG-EVANS: OESTROUS CYCLE IN THE RAT 83 to be further along than that which may be found on the tenth day of pregnancy (fig. 64, pi. IAr) when, as we have already stated, these changes begin in this area of the vagina, thenceforth increasing and advancing downward in the viscus (Rat 3587). It is apparent from the above that infertile copulation or cervical stimu- lation brings on conditions in the vagina identical with those occurring in the early days of pregnancy. The condition has consequently been called, by us “pseudo-pregnancy,” a term which has already been employed by Hill and O ’Donoghue to characterize the sequelae of normal ovulation in the marsupials. It would be highly interesting to find out whether in these primtive mammals coitus with vasectomized males might prolong and possibly heighten the con- spicuous changes in pseudo-pregnancy. We may, however, at once admit that we have not been able to obtain these effects in the guinea pig, where oestrus is not delayed by infertile copulation, and so would refer again to the conjecture previously made in this paper that this remarkable effect of coitus in the rat is either peculiar to this animal or to those which like it have an oestrous cycle so short that the tubal journey undertaken by the fertilized ova would not he com- pleted and the uterus reached before another spontaneous oestrus would occur to prevent their implantation and thus destroy them. Effect upon the ovary. Studies on the ovaries of animals which have been submitted to cervical stimulation or coitus with vasectomized males show the persistence and con- tinued slow growth of the corpora lutea produced at the oestrus when the ex- periment was inaugurated, i.e., the suspension of oestrus behavior on the part of the animal and of oestrous changes in the vaginal mucosa is correlated with the suspension of ovulation and the continued function of the last formed cor- pora lutea. These structures do not increase in size so rapidly as do the corpora lutea of gestation or lactation, hut nevertheless they do attain diameters of from 1.2 to 1.4 millimeters. During most of the prolonged dioestrous interval following coitus or cervical stimulation the corpora retain a lipoid content identical with that of the healthy corpora of ovulation or pregnancy (fig. 87, pi. VII, and table 18, p. 64). At the end of such an interval the lipoid granules begin to increase somewhat in size and number and resemble the change accom- panying the normal corpora of ovulation at the onset of the next oestrus (tig. 88, pi. VII). Here, again, consequently, we have a clear instance of the final de- generation of corpora lutea preceding the occurrence of oestrus. The proof of the suspension of ovulation both after copulation by vasec- tomized males and after stimulation by mechanical means is easily carried out by the use of vital dye. Dye was administered in several cases during preg- nancy ; soon after parturition the animal was allowed to copulate with a vasec- 84 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA tomized male, and some fifteen days later, when she would copulate again, was killed and the ovaries preserved. In such ovaries are to he found only two sets of corpora, one stained blue (the corpora of pregnancy) and the other colorless and consisting of bodies all of the same character (the corpora of copulation). If ovulation had occurred during the long interval there should have been cor- pora constituting more than one set, distinguishable from each other by differ- ence in lipoid content. Similar conditions are found and the same reasoning applies if dye is given during lactation, the cervix stimulated at the first oestrous stages following weaning and the animal killed at the next oestrus. The structure of what might be called the corpus of copulation or of stimu- lation or pseudo-pregnancy does not differ in its essentials from that of ovula- tion and gestation. (See table 18, p. 64.) It will be seen from this brief account that the corpus of copulation or stimulation is in a sense one of ovulation par- tially transformed into one of pregnancy, the transformation not being com- plete because the forces or stimuli operating late in pregnancy are lacking here Those stimuli to be found in the condition of the uterus as the result of the presence of developing embryos, so far can not be duplicated even by the for- mation of deciduomata as described further on. Consequently the cycle comes to an end before complete transformation is possible. From the nature both of the ovary and the vagina the condition resulting from stimulation of the cervix is justly called “ pseudo-pregnancy. ’ ’ D. Production of Deciduomata Robert Frank was the first to ascertain that the experimental production of deciduomata, first discovered by Leo Loeb in the guinea pig, can be obtained in rats, and Corner and Warren in this laboratory have been able to confirm Frank’s statements. Indeed, it appeared to the latter observers that deciduo- mata, could be produced with peculiar ease in this species inasmuch as practi- cally all their experiments were successful. They were led to assign their success to the constant presence in the ovary of an abundant amount of lutein tissue, using Loeb’s hypothesis of the dependability of deciduoma production upon active corpora. Our experience is not in accord with this, although, as will be shown later on, it can probably be utilized as an explanation of such results. We may state at once that we have not been able to produce deciduo- mata in animals which are experiencing normal four day ovulation cycles. On the other hand, typical tumors of this sort, often of considerable size, may be produced under most conditions which involve a delay in ovulation. ( Figs. 90, 91, pb I; 92 to 95, pi. XI.) The present account has already demonstrated some physiological and some experimental conditions in which such delayed ovulation occurs, viz., after copulation or stimulation of the cervix, during lactation, and in the case of animals which have barely reached maturity. LONG-EVANS: OESTROUS CYCLE IN THE RAT 85 The foreign body uniformly employed in our experiments for deciduomata production consisted of loops of black silk thread inserted in either horn of the uterus in such a way that they passed clearly into the lumen and out again to be tied loosely (fig. 90; compare with fig. 91, pi. I). Although the studies of other investigators have indicated that so simple a thing as trauma alone would be adequate to elicit a deciduomatous response, the thread has the further ad- vantage of leaving one in no uncertainty as to the exact site of uterine injury, an advantage of some consequence in enabling one to be sure that a completely negative response has ensued when the specimen is examined microscopically. As an answer to the objection that so grave a thing as the administration of an anaesthetic and the opening of the body cavity with handling of and per- haps a trauma of the abdominal viscera would itself delay ovulation and give us an abnormal cycle, we may state that our operations, which were carried out with all of the care and celerity possible, showed that, when so conducted, the procedures themselves seldom interrupt appreciably the onset of the next oestrus. Irritation of the uterine mucosa by injury or by the presence of threads does not affect the length of the oestrous cycle. This is in striking contrast to the profound effect produced by the simple stimulation of the cervix. It would seem to mean that the portion of the uterus that will respond in this way is very narrowly localized to the part normally affected by the vaginal plug. 1. DURING NORMAL OVULATION CYCLES In twelve cases at various times during the oestrous cycle we placed silk threads in either horn of the uterus, examining the organs again at the time of onset of the next oestrus. In no case was any deciduoma-like tumor produced (table 27). Further experience has inclined us to attribute two causes for this uniformly negative result. First, we now believe that the uterine epithelium is not sufficiently sensitive to produce a typical deciduomatous response until the cycle is at least three days old. Secondly, if the experiment he carried out this late in a normal ovulation cycle another oestrus will supervene within a day, effectually destroying the freshly forming tumors. That fertilization of the ova at any particular ovulation time leads to the immediate suspension of succeeding oestrous cycles has, of course, been known for a long time, but it might be supposed that this rule would not hold in animals like the rat which have so short an oestrous cycle that it is less than the time required for the transit of the fertilized ovum through the Fallopian tube and the establishment of uterine implantation. Data of other observers (cf. Huber, et al.) indicate that the developing ova of the rat reach the uterine lumen about four days after copulation and remain a day or two before implantation. It is 86 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA thus conceivable that in the rat another oestrous cycle could occur before actual implantation of the developing embryo, were the occurrence of implantation the only cause for withholding the next oestrous cycle after fertilization, but we have shown that copulation withholds the next oestrus. Now this effect of copulation seen in the rat may itself be unique and may with good propriety be looked upon as necessary to the successful implantation of the fertilized and developing ovum, for, as we shall show later, there is good reason for believing that if oestrus did occur the uterine degenerative changes would effectually prevent implantation or bring to decay and resorption the freshly implanted eggs. Operations for the production of placentomata during the normal cycle. Table 27. Designa- tion of animal Operated upon during Killed days after operation 3539 Stage 1-2 3 Killed in long Stage 3. No swelling; few decidual cells. 3596 Stage 2 4 Negative. Killed in next Stage 2. Uterus distended. 3530 Stage 4 8 Negative. Killed in next Stage 2. 3536 Stage 4 (end) 8 Negative. Killed in next Stage 4. 3581 Stage 4 (end) 4 Negative. Killed in next Stage 1-2. Uterus distended with fluid. 3523 12 hrs. after Stage 4 6 Negative. Killed in next Stage 4. 3528 First 24 hrs. of interval 8 Negative. Killed in next Stage 2. 3533 First 24 hrs. of interval 5 Negative. Killed in early next in- terval. 3584 First 48 hrs. after Stage 4 4 Negative. Killed in early Stage 4. 4026 3 days after Stage 2 7 Negative. Observed 2 days after op- eration. No cycle. 3617 2 days post- partum 4 Negative. Cycle not known. Left horn larger. 3631 % 5 days post- partum 4 Negative. Killed after new Stage 2. In many animals with a longer cycle the next expected oestrus would occur after implantation were well established. We have found that in at least one of these cases, the guinea pig, copulation itself does not defer oestrus. These facts tit with the conception that in these animals foetal hormones are from the beginning responsible for the suspension of oestrus during pregnancy. It is our belief that while in the inseminated rat the initial inhibition of oestrus is due to cervical stimulation, the continued absence of cycles is here also the eftect of foetal tissue. LONG-EVANS: OESTROUS CYCLE IN THE RAT 87 Table 28. Operations for the production of placentomata during pseudopregnancy. Desi gna- tion of animal Time of stimulation Operation days after Stage 2 Observed days after operation Killed days after operation Findings 3585 Stage 2 2 4 Slight enlargements. 3611 Stage 2 2 6. New loops 6 after last opr. Negative. No cycles. 3614 Stage 2 2 6. New loops 6 after last opr. Negative. No cycles. 3634 Stage 2 2 6. New loops 6 after last opr. Negative. No cycle. 3732 Stage 2 2 6 Negative. Killed after new Stage 2. 4077 Stage 2 3 6 Negative. No cycle. 3630 Stage 2 4 2 Very slight swelling. Few decidual cells. 3636 Stage 2 4 2 Very slight enlargement. Few decidual cells. 3597 Stage 2 4 4 Moderate enlargements. 3639 Stage 2 4 6 Great enlargements. 7 mm. 3640 Stage 2 4 6 Great enlargements. 6 mm. 4447 Stage 2 4 6 Negative. No cycle. 4230 Stage 2 6 4 Negative. No cycle. 3613 Stage 2 6 4 Negative. Killed on day of new Stage 1-2. 3578 Stage 1 6 4 Negative. Killed on Stage 2. 3641 Stage 2 6 5 Negative. Killed on new Stage 2. 3642 Stage 2 6 5 Negative. Killed on new Stage 1-2. 4117 Stage 2 6 6 Negative. No cycle. 4076 Stage 2 6 6 Negative. No cycle. 3729 Stage 2 7. Large swelling 4 after ob- servation Negative. Killed on new Stage 2. 3730 Stage 2 6 7. Cornified 4 after ob- servation Negative. Killed after new Stage 2. 3731 Stage 2 6 7. Cornified 4 after ob- servation Negative. 54 Stage 2 6 7 Negative. No cycle. 4041 Stage 2 8 2 Negative. No cycle. 4257 Stage 2 8 21 Negative Two sets of oestrous stages elapsed after operation. Cervix stim- ulated at last Stage One. Killed 6 days later. 3576 12 hrs. post part. 4 4 Negative. 3598 18 hrs. post part. 2 5 Negative. 3624 4 and 9 hrs. post part. 4J4 hrs. 4 Negative. 3447 Stage 2 4 after stim. 6 Negative. Ovaries removed at oper. 3548 Stage 1-2 4 after stim. 6 Negative. Ovaries removed at oper. 88 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA 2. DURING THE PAUSE FOLLOWING CERVICAL STIMULATION ('‘PSEUDO-PREGNANCY”) Inasmuch as our studies had taught us that after coitus or stimulation of the cervical canal the reproductive organs are in a condition resembling that of pregnancy, it was apparent that we had conditions which would enable us to test whether or not the response of the uterine epithelium to a foreign body was superior in character to that which we had been able to discover during normal cycles. Accordingly, at Stages One and Two a slender glass rod was introduced into the cervical canal of some thirty animals and at various times thereafter silk threads passed through the uterine walls and lumen and tied in position, the animals being sacrificed at various time intervals thereafter. Table 28 shows the result of these operations, which, in brief, is that if the operation is done on the fourth day after cervical stimulation and the animal sacrificed from the fourth to seventh day thereafter, providing no oestrus has recurred, excellent deciduomatous tumors well over a centimeter in diameter may be produced. In no case after the incidence of the next oestrus is a tumor of this sort ever found, even though the time of operation and autopsy would otherwise lead one to expect that the response had taken place. We were therefore very defi- nitely of the impression that the changes in the uterus associated with oestrus evidently brought these tumors to sudden atrophy and resorption, and we were confirmed in this interpretation by an experiment in which we were actually able to satisfy ourselves by laparotomy and inspection that typical deciduo- mata were actually present just before the onset of oestrus. The animal was closed up and killed at the incidence of the next oestrus. The uterus showed no evidence of the previously prominent uterine tumors and the sections no deciduomatous tissue (Rat 3729). It is interesting that a thread after being placed in the uterus in one cycle never calls forth a response in the next cycle. A number of cases (not listed) prove this. The explanation would seem to lie in the fact that the threads be- come covered by a growth of cells. It has been found that in tissue cultures silk forms an excellent substratum along which cells may creep. Such a cell covering perhaps protects the surrounding tissue from the foreign body and hence from further irritation. Applying this idea to the failures following early operation, the suggestion presents itself that in early operation the threads became covered and so were prevented from exerting an influence at the ap- propriate time a little later. LONG-EVANS: OESTROUS CYCLE IN THE RAT 89 3. DURING LACTATION Under natural conditions the greatest delay in ovulation is, of course, that occasioned by the period of lactation which may not infrequently be prolonged to thirty days. It hence occurred to us that ideal conditions for the production and growth of deciduomata must exist during lactation. Here is a span of time greater than any other, even including the period of gestation, in which the ovary had been proved by us to refrain from discharge of ova and to nourish a set of large corpora lutea. We consequently placed silk loops in the uterus of twenty rats at various time intervals after the day of littering and while they were nursing normal sized, healthy litters, sacrificed them at various time intervals thereafter, as shown in table 29. Table 29. Operation for the production of placentomata during lactation Designation of animal Operation days postpartum Killed days after operation Number of young suckled Findings 3638 4 4 3-9 Some evidence of placentomata having degener- ated. Suckling irregular. 4680 4 6 6 Large swellings. (5-6 mm.) 4671 4 6 6 Large swellings. (5 mm.) 4670 4 8 9 Large swellings. (5-6 mm.) 4681 4 10 6 Negative. 69 4 11 7 Negative. 59 5 11 5 Negative. No decidual cells in sections. 3637 6 4 8 Large swellings (4.25 mm.) 3591 7 4 8 Definite enlargements. (2.5 mm.) 3620 8 4 5 + Definite enlargements. (3 mm.) 3623 10 4 Large swellings. (4-5 mm.) 4678 10 8 4 Large swellings. (6 mm.) 4691 10 6 7 Negative, but left horn distended in region of threads. 4679 10 6 3 Negative; left horn slightly larger. 4672 11 6 5 Small swellings, distinct. (3-4 mm.) 4724 16 5 Slight swelling at threads. (2.5 mm.) 3703 16 6 4 Decided swellings at threads. (3.5 mm.) 4674 16 6 6 Slight enlargement. 58 5 11 Negative. Ovaries removed at operation. 61 4 11 Negative. Ovaries removed at operation. Inasmuch as our operations upon the uteri of animals after stimulation of the cervical canal had demonstrated that the deciduomatous response is greater when the cycles are three or four days of age, we picked upon this as the earliest time at which the silk loops could he advantageously inserted in the uteri of lactating animals and expected that a long continued growth might be obtained of the deciduomata thus produced, since lactation serves to maintain for a month a constant condition of affairs in the reproductive tract. Three experi- ments were consequently done in this manner (Rats 4681, 69 and 59) in which 90 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA a wait of ten or eleven days took place before autopsy, but to our surprise no uterine tumors were then present. On the other hand, three similar experi- ments in which a wait of only six or eight days was enjoined gave large and histologically typical tumors (Rats 4680 [fig. 90, compare with fig. 91, pi. I], 4671 and 4670). It appeared probable that in the former three cases the deci- duomata after having been produced had gradually degenerated. It was ap- parent that an irritant like the silk thread could provoke the production of these tumors when the uterus is in the proper “ sensitive ” state, but that these tumors had a limited span of life after their formation and that some other factors were necessary for their continued maintenance and growth. It next appeared important to us to discover how late in lactation the uterus was “sensitive,” providing due attention were paid to the fact that the four or five day interval after the threads were inserted is the ideal time to discover deciduomata in full vigor. Operations at six, seven, eight, and even ten days post-partum gave typical deciduomatous uterine responses when the animals were sacrificed four days after the operation (Rats 3687, 3591, 3620, 3623). Indeed, in one case when a wait of eight days was enjoined before autopsy, large typical tumors were still found (Rat 4678). Although the cases of operations conducted at a still later time during lactation are too few in number, we may state that operations even at the sixteenth day post-partum may succeed in producing a characeristic tumor (Rat 3730). To summarize: The evidence presented in tables 27, 28, and 29, in spite of a number of individual discrepancies, shows that the experimental production of deciduomata in the rat might be viewed as dependent upon the presence of healthy, functional corpora lutea at least three days of age and that all condi- tions in which the life of corpora is prolonged serve also to prolong the time during which these tumors can be provoked by irritation of the uterine epithe- lium. Most of the discrepancies occurring in tables 27 and 28 are explainable by the appearance of oestrus before the experiment was terminated and by our clear demonstration that oestrus is inimical to growth, or even maintenance, of these tumors. The evidence of the production of deciduomata is thus strictly consonant with the view that normal implantation is brought about through the aid of lutein cell hormones, hormones which are apparently produced only by corpora that have attained an age corresponding to the time interval normally occupied by the transit of fertilized ova through the Fallopian tube—in the rat an interval of about four days. 4. EFFECT OF OVARIOTOMY Corner and Warren in this laboratory have previously established the fact that after ovariotomy deciduomatous tumors can never be provoked by uterine irritation, and four cases studied by us (Rats 3447, 3548, 58, and 61) further confirm these findings. LONG-EVANS: OESTROUS CYCLE IN THE RAT 91 VIII. EFFECT OF ABLATION AND TRANSPLANTATION OF VARIOUS PORTIONS OF THE REPRODUCTIVE TRACT UPON THE NORMAL OESTROUS CYCLE, PREGNANCY, AND LACTATION A. Hysterectomy Vague surmises have existed from time to time in the gynecological literature as to a supposed detrimental effect on the ovaries of complete removal of the uterus, the contention being occasionally made that the menopause is hastened by such a procedure. It consequently appeared to us of some importance to examine the effect on the oestrous cycle of complete ablation of both horns of the uterus. Inasmuch as in such animals as the rat and guinea pig the cyclic phenomena in the reproductive tract manifest themselves in the vagina as well as in the uterus, it should he possible to detect in such a mammal whether or not hysterectomy destroys this periodicity. Such evidence would he more diffi- cult to obtain in man where cyclic changes in the vaginal fluid are not known independently of menstruation, which is occasioned by uterine hemorrhage. Accordingly, in vigorous young females possessing normal oestrous cycles the uterus was completely removed save for a small portion of the cervix. The vaginal smear was examined daily for several months thereafter, and in no instances, except for an occasional short pause immediately following the oper- ation, was there any interference with the regular progress of normal oestrous cycles (table 30). These experiments would thus appear to demonstrate con- clusively that hormones from the uterine tract are not concerned in the causa- tion of the cyclic phenomena of oestrus, which, in the absence of this viscus, run their usual course and provoke their usual changes in the mucosa, even of the more external portion of the generative tract—the vagina. The typical effect of cervical stimulation in causing the delay of ovulation was also secured in these hysterectomized animals. The importance of the cervical canal in the procreative act of this animal led us to desire to test, however, whether this tissue had a more important par- ticipation in oestrus than the uterus. Accordingly, in several cases we at- tempted a complete excision of the cervix and upper portion of the vagina along with both uterine horns. The operation is not easy, one difficulty being in the preservation of the ureters. At least a single entirely successful case of such an ablation was accomplished and the daily vaginal smear for many weeks thereafter showed the regular recurrence of normal oestrus. 92 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Hysterectomy is equally innocuous in its effect upon lactation, a result which we might conceivably have been led to expect inasmuch as ovariotomy itself does not gravely influence lactation. Several cases of the combined excision of ovaries and uterus during lactation have given similar results, the litters evi- dently enjoying normal nutrition, to judge from their vigorous growth. Table 30. Length of oestrous cycles after hysterectomy. Designation of animal Length of cycle in days 3579 11, 4, 4, M14* 4, 4, 4, 4, 4, M5, 4, M4, M16, 4, 11. 3599 14, 5, 4, 4, M19, 4, 4, 4, 4, M21, 5. 3820 11, 4, 4, 5, 6, M14, 4, 5, 5, 4, M13, M20, 12, 4, 7, 8, no more cycles (animal very thin). 4869 10, 4, 5, 5, 4, 5, 8, M8, 6, 4, 4, M4, M25, 4, 5. 4870 17, 6, 4, 4, M15, 4, 4, 4, 4, 4, M15, 4, 4, M14, 5. 4871 Died 10 days after operation; no cycles. 4872 16, 4, 5, 4, M15, 4, 5, 4, 4, 4, M16, 4, 4, M13, 4. 4873 13, 12, 17, 4, 4, 4, 4, 5, 5, M20, 6, M17. 4874 15, 6, 4, 4, M20, 5, 4, 4, M4, M18, 7, cervix removed, 8, 6, 12, 9, died. 4875 3, 5, 3, 5, 6, 4, 4, 4, M13, 5, 4, 5, 4, M17, 4, 4, M8, 5, 4. 4876 14, 8, 9, 6, 4, 6, 21, 4, 7, 7, 3, cervix removed, 4, died. 4878 Ovaries and uterus removed; suckled for at least 11 days. *M cycle preceded by mechanical stimulation of cervix B. Excision of Mammary Glands On account of the extensive, irregular, and somewhat ill-defined tract occu- pied by the mammary glands of the adult, the total excision of the mammary apparatus in adults is difficult, if not impossible. Within the first eight days after birth the mammary anlagen are, however, confined to an area within one or two millimeters of the nipple and their extirpation can be easily carried out under the binocular microscope. The twelve small cutaneous defects created by the operation do not require suture, the young epidermis closing over with surprising speed. We experienced no difficulty in getting such young to be received again by the mother and to nurse normally. The animals mature as do normal ones, nor is the oestrous cycle of such animals in any way atypical (table 31). The participation of the mammary apparatus in the oestrous cycle, which has come to be a well established fact, would thus not appear to be an essential link in the chain of functional factors in oestrus. It is evident that the underlying cause for oestrus affects the mammary just as it does the uterine or vaginal tissue. LONG-EVANS: OESTROUS CYCLE IN THE RAT 93 Nor can it be urged that these glands, characteristic of the mammalia, exert any essential effect during pregnancy, although profound changes in them occur during the latter event. Five individuals in which the mammary apparatus was completely ablated were allowed to mate, the resulting pregnancy being normal in duration and all other observable respects. Table 31. Maturity and oestrous cycles in rats after total excision of the mammary glands during infancy Designation of animal Age at operation Age at maturity Length of successive cycles in days 3746 8 days *78 days 8, 8, 5, 5, 5, 7, 7, fM13, 4, 4, M10, (712, <712, 4, 4, 4, 5, 6, 4, 4, 4, 4, 4, 4, 4, 4, 4, M12, 4, Mil, M13, 4, 4, 4, 4, 4, M10, 4, Mil, 4, 4, 4, 4, 4. 3747 8 days *78 days 13, 6, 5, 6, 4, 7, 3mm. M10, 9, M10, 6, 6, M10, 4, c? preg. 22, 4, 5, 7, 4, 4, 4, 6, 4, 5, 4, 6, M12, 5, 5, M14, 4, 4, M12, Mil, 4, M13, 4, 4, 4, 4. 3748 8 days *79 days 30f before first cycle, 5, 4, 3 mm. M8, 3, M12, 5, 4, 11, 4, M15, <715, 5, 7, 7, 4, 5, 4, 4, 4, 4, 3, 4, 4, M13, 5. 3749 8 days 102 days 4, 4, 5, 4, 15, 14, 5, 4, 9, M4, 10, 4, 4, <7 preg. 24, 3, 4, 4, 5, 4, 4, 4, 4, 8, 4, 5, 4, Mil, 4, 5, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4. 3750 8 days *78 days 12, 5, 5, 4, 5, 9, 8, 5, M8, 6, M4, 9, M3, <715, 4, 4, 713, 3, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, M16, 4, M5, 8, M3, M13, 3, 4, 4, M13, 4, 4, 5, M4, 4, 4, 4, 4. 3751 3 days *73 days 12, 4, 5, 8, 6, 7, 3, FM15, 4, 4, FMll, 4, 7preg. 22, 12, 5, 4, 5, 4, 5, 4, 4, 4, 4, 4, M14, 4, M5, 713, 4, 715, 5, M12, 5, 4, 4, 4, 4, 4, 4,5 . 3752 3 days 74 days 21, 7, 7, 16, 5, 11, no cycles Nov. 3 to Feb. 11, sickly. 3753 3 days 78 days 18, 13, M12, 5, 11, 4, 5, 4, M5, 4, 713, 713, 7preg. and resorption, 35, 4, 4, 4, 4, 4, M4, M4, 5, M4, M4, 3, 5, 4, M4, 4, M4, M16, 5, 5, 4, M4, 4, 4, 4, 4, 4, 4. 3754 3 days 86 days 9 days before first cycle, 9, 6, 6, 11, M9, 5, 5, M5, 7, M5, 5, 712, 711, 5, 5, 7, 5, 4, 4, 4, 4, 4, 4, 4, 4, M6, 4, 4, 4, M14, 8, 8, 10, 5, 11, 8, 7, 4, 4, 4. 3755 3 days 94 days 8, 10, 8, M4, M5, 6, 6, 4, 6, 5, FM14, 715, 4, 5, 5, 9, 5, 5, 4, 4, 4, 5, 3, 5, 3, 4, 5, 8, 4, M4, 4, 9, 8, 9, 5, 5, 4, 5, 5, 5, 4, 12. 3756 3 days 77 days 18, 6, 12, 15, 12, 7, FM5, 4, FM5, 4, 4, 7preg. 23, 14, 11, 5, 5, 4, 4, 4, 8, 4, 4, 4, M13, 3, 6, 10, 13, 4, 4, 4, 5, 4, 5, 4, 4, 4, 4. *Age when observations were begun—vagina open. M. Indicates mechanical stimulation of cervix by insertion of a rod; figure following, the length of next cycle. FM. Failure to introduce rod into cervix. 3mm.M. Rod introduced only 3 mm. In other instances inserted about 10 mm. cf Copulation with normal male. 94 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA C. Ovariotomy We can confirm the experience, which is universal, that excision of the gonads brings about a sudden and permanent cessation of all cyclic changes in the reproductive tract. A double ovariotomy was carried out upon twenty-five rats at various times in the normal oestrous cycle. In some cases, notably those in which the operation was carried out immediately preceding an expected oestrus, the next expected typical oestrous changes in the vaginal smear occurred so that it is evident that the impulse or cause of oestrus operates slightly an- terior to the actual event. In all cases of ovariotomy the vaginal smear showed henceforth no indications whatever of an oestrous cycle, daily observations hav- ing been carried out for some months subsequent to the operation. The effect of extirpation of the gonads upon the structure of the uterus and vagina, where a gradual atrophy is encountered, has already been reported hv Marshall. We are able to confirm the discovery that ovarian ablation interferes in no respect with the continuance of lactation, vigorous litters having been reared after such ovariotomies (Rats 53 and 54). Previous observers (Frankel, 1903, for the rabbit; Marshall and Jolly, 1904, for the rabbit and rat) have reported the necessity of the ovary (on account of its possession of lutein tissue) for the implantation and nutrition of the early embryo, hut have indicated that the ovaries may he ablated in the latter half of the pregnancy with impunity. D. Ovarian Transplantation The detection of the oestrous cycle in the rat by means of the vaginal smear and the fact that oestrus never recurs after ovariotomy, furnishes us with an ideal test for the success of ovarian transplants. We have carried out trans- plantation of the ovaries in twenty-nine cases, the site of the graft being in the mesometrium, omentum, spleen, or rectus muscles. In all cases vigorous young females were employed. Nineteen of the cases were instances of autotrans- plantation (table 32) where the expectation of success is, of course, greatest, and we obtained fourteen favorable results here as against five failures. Ten cases were attempted of homotransplantations (table 33) and in such opera- tions both the donor and recipient were prepared simultaneously so that the exchanges were effected in the shortest possible time. It is to be noted that only one entirely successful outcome attended these efforts, seven of the cases having given at no time indications of the success of the graft and two of them a single oestrous cycle only. In the two cases in point, the single post-operative oestrus produced occurred so late (seven days) that we are not to attribute this LONG-EVANS: OESTROUS CYCLE IN THE RAT 95 to the spontaneous occurrence of the next expected oestrus, as sometimes hap- pens after ovariotomy, but must admit an ephemeral “take” of the graft with secondary and speedy resorption. Doubtless more success would attend homo- transplants which were between mother and daughter possessing tlig «amA m-tlo In all instances the ovary to be transplanted was cut into twro or more pieces to facilitate vascularization before decay, which in the case of large pieces of tissue may be extensive. Daily observations of the vaginal smear have been carried out for somewhat over seven months subsequent to these operations. They demonstrate that the successful ovarian grafts produce the next oestrous cycle as a rule one week after the operation, this varying in the time of its occurrence from five to twelve days after the transplants were made. In practically all the successful cases throughout this long time interval (seven months) a regular succession of typ- ical oestrous cycles of normal length lias occurred. When observations were discontinued the animals were almost one year of age, which approaches the end of the normal period of full reproductive vigor for this species. Our ex- periments thus do not lend any support to the assumption that the life of the ovarian transplants need be shorter than that enjoyed by the gland in its normal situation. The oestrous changes experienced by animals with transplanted ovaries were, as far as we could determine, entirely normal. In order to discover whether cyclic changes in the transplanted ovaries corresponded with oestrous stages in the vagina we examined the transplants of several typical successful cases at the stage in the vaginal smear when leucocytes first appear among the cornified epithelial cells, a stage in which under normal conditions a fresh ovu- lation should have occurred. In such instances the transplanted ovary (fig. 96, pi. IX) disclosed a group of new corpora together with other corpora from the immediately preceding ovulations at four-day intervals. A normal content of healthy follicles and of follicular atresia was also present in these ovaries so that study of them did not indicate any departure from the normal phenomena found in this gland. The transplanted ovaries evidently mature follicles which rup- ture at regular intervals and are transformed into corpora. Under the circum- stances the ova are evidently expelled into the tissue immediately surrounding the transplant rather than, as normally, into the fluid-filled periovarian space. They evidently speedily degenerate, for in two instances where the ovulation must have been recent no traces of eggs were discovered. Difficulty in the expulsion of the ovum under these conditions might explain the occurrence of several cases of the retained egg about which the corpora have formed, an oc- currence by no means absent in the normal ovary, as we have already shown, but where the proportion of such cases is, of course, somewhat lower than we would expect in these transplanted cases. 96 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA Table 32. Effect of ovarian autotransplantations upon the oestrous cycle as determined by vaginal smears. Designation of animal Position of transplant Interval be- tween opera- tion and first cycle in days Length of cycles in days 3883 Spleen 10 days 8, 6, 5, M13, 19, 11, 10, 3, 5, 2, 5, 4, 6, M17, 6, M15, 4, M4, M14, 5, 4, M4, 8, 6, 5, 5. 3885 3886 Spleen Spleen 8 days No cycles. 5, M4, 4, M4, 4, 17, 5, 4, 8, 3, 5, 7, 3, 5, 5, 4, 8. 3887 Spleen 12 days 8, 8, 3, 3, 4, 8, 4, 4, 6, 4, 4, 6, 4, 4, 8, 4, 4, 7, M13, M13, 4, M6, 7, 11, 15, 7. 3888 Spleen 7 days 5, 5, M6, 4, 5, 5, 6, 4, 7, 5, 4, 5, 6, 6, 4, 4, 4, 5, 6, 6, 6, M20, 5, 3, M4. 3889 Spleen No cycles. 3890 Spleen 5 days 5, dM, 4, 5, 6, M4, M12, 4, 4, 4, 4, 4, 4, 4, 3, 4, 5, 4, 5, 4, 4, 4, M4. 3891 Spleen No cycles. 3892 Spleen No cycles. 3893 Spleen 7 days 6, 4, MdH, M9, 16, 4, 4, 4, 5, 5, 6, 4, 4, 4, 4, 4, 4, 7, M18, 5, 5, M15, 5, M17, 5, 5, 4, 6, 13, 5. 3894 Spleen No cycles. 3895 Spleen 12 days 9, 5, 6, M10, 10, 43 days without cycles. 3896 Rectus muscle 7 days 6, M3, M and