THE DEVELOPMENT OP THE CORPUS LUTEUM. 189

The Development of the Corpus Luteum:a Review.

ByFrancis H. A. Marshall, ITI.A., D.Sc,

Carnegie Fellow, University of Edinburgh.(From the Physiological Laboratory of the University of Edinburgh.)

THE present paper is the result of an attempt to collecttogether and give an account of the literature of the for-mation of the corpus luteuni, during the last ten years, thatis, since the publication of Sobotta's first paper on the corpusluteum of the mouse.

Of the three original hypotheses pub forward to explain themode of formation of the corpus luteum, and the origin of thelutein cells, that of Paterson, who regarded the structure asderived from the blood coagulum left in the cavity of theGraafian follicle after its discharge, gained few or no ad-herents among subsequent investigators. The other twotheories, those of von Baer and Bischoff, on the other hand,have each received a considerable amount of support. VonBaer supposed the corpus luteum to be a connective-tissuestructure, in the formation of which the membrana granulosaor follicular epithelium had no share; while Bischoff con-cluded that the lutein cells were formed by the hypertrophyof the epithelial cells of the undischarged follicle. Amongthe principal supporters of von Baer's view appear the namesof Leuckart, His, Kolliker, Slavjansky, Gegenbaur, Benckiser,Schottlander, and Minot. Those who have adopted thealternative theory of Bischoff include Pfliiger, Waldeyer, Calland Exner, Beigel, and Schulin.

190 FBANOIS H. A. MARSHALL.

The first really systematic effort to deal with the questionwas made by Sobotta, whose eai-liest paper on the subject waspublished in the ' Anatomischer Auzeiger' in 1895. In thefollowing year the complete paper was issued. These papersdescribe an investigation on the development of the corpusluteum in the mouse, in which a large series of stages wasexamined, each of them being collected according to a definiteplan, the animals being killed at known intervals after theoccurrence of coition, in reference to which the period ofovulation had been previously determined. Sobotta's in-vestigation resulted in confirming Bischoff's view that thelntein cells are the much hypertrophied epithelial cells of theundischarged follicle, the connective-tissue element whichforms an anastomosis among the lutein cells being derivedfrom the inner layer of the theca. The theca esterna isdescribed as taking no share in the ingrowth, while the thecainterna is stated to become entirely used up in the formationof the interepithelial network. The hypertrophy of theepithelial cells is described as being of the nature of asimple enlargement, unaccompanied by cell division. Thecavity of the follicle is said to become eventually filled up bya central plug of connective tissue.

The conclusions reached by Sobotta regarding the for-mation of the corpus luteum were subsequently corroboratedby him in an investigation on the corpus luteum of the rabbit,the stages of development being also obtained by killing theanimals at stated intervals after coition.

In 1898 Stratz published descriptions of stages in. the for-mation of the corpus luteum of Tars ius , Tupaia, andSo r e x ; and these agree in all essential particulars with theaccounts given by Sobotta.

The development of the rabbit's corpus luteum was alsostudied by Honore, who adopted the same method as thatemployed by Sobotta. According to Honore the inter-epithelial proliferation of connective tissue is derived in partfrom the theca externa, and not exclusively from the innertheca, as supposed by Sobotta; while the theca interna is

THE DEVELOPMENT OF THE CORPUS LUTEUM. . 191

stated to be not entirely exhausted by the ingrowth, some.part remaining to form a layer within the outer theca, afterthe.full formation of the corpus luteum.

Kreis's observations on the young human corpus Inteumlikewise support the hypothesis put forward by Bischoff.

.Belloy, who investigated the formation of the corpus luteumin the rat and guinea-pig, while regarding the lutein cells as

• being derived from the follicular epithelium, describes anactive proliferation of these cells soon after the follicle'srupture. No figures are given by Belloy, and it seemspossible that he has confused the ingrowing cells of con-nective tissue from the theca interna with the membranagranulosa cells. Bouin, who also investigated the corpusluteum of the ra.t and guinea-pig, reached conclusions similarto those of Belloy.

Heape, without entei'ing into a discussion on the origin ofthe lutein cells, lays some stress on the absence of diyisionamong these cells in the ovaries of certain monkeys, pointing

-out that the enlargement is the result of a simple hyper-trophy.

Babl, writing especially on the human corpus luteum, con-cludes that the lutein cells have a double origin, arising bothfrom the membrana granulosa and from the theca interna.

A number of investigators, on the other hand, since thepublication o£ Sobotta's work, have adopted the theoryoi'iginally put forward by von Baer, that the lutein cells arisefrom the connective-tissue wall, the follicular epitheliumbeing either completely discharged along with the ovum andthe greater part of the liquor folliculi, or else being partiallydischarged and partially degenerating in situ. Among thoseholding this view are His, Kolliker, and Paladino, who havelately reiterated their former opinions.

Von Baer's theory has also received considerable supportin recent years from Nagel, who has described the corpusluteum in the human subject as an entirely connective-tissuestructure. In this he has been followed by Clark, who worked

.on the formation of the corpus luteum in the sow. and in .the

192 FJUNG1S H. A. MARSHALL.

human female, and claimed that the result of his investigationhad put the matter almost beyond question. Clark's accounthas been confirmed by Doering, who also worked upon thesow's corpus luteum. Others who have adopted the viewthat the luteiu cells have a connective-tissue origin areBiihler, Wendeler, and Stockel, who have examined anddescribed developing human corpora lutea.

None of these investigators, however, appear to liave givenan account of the growing corpus luteum in all stages ofdevelopment, while in the case of several of the accounts, it isnot clear that the structures described were not in realityatretic follicles, that is to say, follicles which had undergonedegenerative changes without discharging their ova. On theother hand, the words used in a description by Clark point tothe conclusion that this author was dealing with the degenerateepithelial cells of an atretic follicle. It seems not improbablethat the young human " corpus luteum" which Doeringdescribes was also an undischarged atretic follicle; whileKolliker's opinion that the corpus luteum is a connective-tissue structure appears to he founded on the assumptionthat the changes undergone by discharged follicles andretrogressive undischarged or atretic follicles are identical incharacter. His, and also Biihler, with reference especially toSobotta's work on the mouse, have remarked that it canscarcely be au accidental circumstance that the accountsgiven of the development of the corpus luteum in the largeranimals and in man are radically different from those describedfor the smaller species. That the discrepancy between theaccounts of various investigators depends upon the size ofthe animals employed does not seem, on the face of it, a veryprobable suggestion. It is to be noted further that in theinvestigations of all these writers who have upheld theconnective-tissue theory the ages of the developing corporalutea were unknown, the material in no case being obtainedby Sobotta's method of killing the animals at definite intervalsafter coition.

In 1901 the present writer published a preliminary account

THE DEVELOPMENT OF THE CORPOS LUTEUM. 193

of an experimental inquiry upon the formation of the corpusluteum in the sheep. In this inquiry the sheep were killedat stated periods either after coition or after the animals hadbeen observed to undergo oestrus. The relation which wasfound to exist between the condition of development of thecorpus luteum and the leDgth of the interval that was allowedto elapse between cestrus and the killing of the animalj was initself a strong presumption that ovulation in the sheep occursnormally during oestrus. Thus the approximate age of theyoung corpus luteum or discharged follicle could in everycase be determined. The result of this investigation was toconfirm in all essential particulars Bischofi's theory, whichhad been accepted by Sobotta. The sheep, however, wasfound to present some differences from the mouse in regardto the mode of formation of the corpus luteum, the connectivetissue ingrowth beingderived partly from the theca externa, andnot merely from the theca interna, and the follicular epitheliumcontinuing to undergo division after the rupture of the follicle,but with greatly decreased frequency. The former of thesetwo observations is in agreement with Honore's statement inregard to the interepithelial connective tissue in the rabbit.The theca interna was said to become entirely used up in theformation of the connective-tissue ingrowth, this statementagreeing with Sobotta's description, but differing from thatof Honore. Two years later the complete account of thedevelopment of the corpus luteum in the sheep was published.

The description given in these papers is thus completelyopposed to His's suggestion that the mode of formation ofthe corpus luteum in the larger mammals is different fromwhat it is in small animals like the mouse and rabbit, unless,as Sobotta remarks, it was intended to include only the ele-phant and the whale in the former category.

Meanwhile, in 1901, the same year in which the preliminaryaccount referred to above was issued, van der Stricht pub-lished descriptions of the developing corpus luteum of batsbelonging to the genera Vespe rugo , V e s p e r t i l i o , andPlaco tus . This author's researches also resulted in eon-