2.C\ 
 
CORNELL UNIVERSITY. 
 
 THE 
 
 Bosttiell ^. Wlomnv IDibirarg 
 
 THE GIFT OF 
 
 ROSWELL P FLOWER 
 
 FOR THE USE OF 
 
 THE N. Y. STATE VETERINARY COLLEGE. 
 
 1897 
 
 Digitized by IVIicrosoft® 
 
Cornell University Library 
 QP 261.M36 
 
 The oestrous cycle and the formation of 
 
 3 1924 000 329 601 
 
 Digitized by IVIicrosoft® 
 
This book was digitized by Microsoft Corporation in 
 
 cooperation witli Cornell University Libraries, 2007. 
 
 You may use and print this copy in limited quantity 
 
 for your personal purposes, but may not distribute or 
 
 provide access to it (or modified or partial versions of it) 
 
 for revenue-generating or other commercial purposes. 
 
 Digitized by Microsoft® 
 
PHILOSOPHICAL TRANSACTIONS 
 
 OF THK 
 
 I 
 
 ROYAL SOCIETY OF LONDON. 
 
 Series Bv VOL. 196, pp. 47-97. 
 [Plates 7-10.] 
 
 THE (ESTROUS CYCLE AND THE FORMATION OF THE 
 CORPUS LUTEUM IN THE SHEEP. ' 
 
 BY 
 
 FEANCIS H. A. MARSHALL, EA. 
 
 LONDON: 
 
 PUBLISHED FOR THE ROYAL SOCIETY 
 
 BY DULAU AND CO., 37, SOHO SQUARE, W. 
 CONTINENTAL AGENTS, MESSES. FEIEDLANDER AND SON, BERLIN. ' 
 
 1903. 
 
 ^B. 216- « ' 10.7.03. 
 
 Price Four Shillings. 
 
 Digitized by Microsoft® 
 
CORNELL UNIVERSITY. 
 
 THE 
 THE GIFT OF 
 
 ROSWELL P. FLOWER 
 
 FOR THE USE OF 
 
 THE N. Y. STATE VETERINARY COLLEGE 
 1897 
 
 Digitized by IVIicrosoft® 
 
Digitized by IVIicrosoft® 
 
Digitized by IVIicrosoft® 
 
III. — The CEstrous Cycle and the Formation of the Coiyus Luieum in the Sheep. 
 
 By Francis H. A. Marshall, B.A. 
 Communicated by Professor J. C. Ewart, F.R.S. 
 
 Eeceived Febnuuy 17,— Read March 12, 1903. 
 
 [Plates 7-lO.J 
 
 Contents. 
 
 Page 
 
 (1) Introduction . . . 4^ 
 
 (2) The (Estrous Cycle 48 
 
 (3) Superficial Phenomena of the Prooestrum and CEstrus ...... 54 
 
 (4) Methods . . 55 
 
 (5) The Histology of the Uterus during the Dioestrous Cycle . .56 
 
 Period of Rest ... ,57 
 
 Period of Growth and Increase of Vessels . . 58 
 
 Period of Breaking down of Vessels and Extravasation of Blood 58 
 
 Period of Recuperation and Pigment Formation .... 60 
 
 Literature of the Prooestrum .... 62 
 
 The CEstrous Cycle in the Ferret .64 
 
 (6) Ovulation ; with Notes on the Atretic Follicle, and the Causes of Barrenness .... 66 
 
 Ovulation in other Mammals . . . .69 
 
 The Atretic Follicle . . .71 
 
 Number of Follicles discharging during CEstrus . . ... . . .73 
 
 (7) The Formation of the Corpus Luteum. . . .... 78 
 
 Other Accounts . 84 
 
 (8) Summary and General Conclusions . . 88 
 
 (9) Efiferences to Literature 92 
 
 (10) Description of the Plates ... 96 
 
 Introduction. 
 
 My investigations on the changes undergone by the genital organs, and more 
 particularly the uterus, on ovulation and its relation to oestrus and coition, and on 
 the mode of development of the corpus luteum in the sheep, were commenced in the 
 summer of 1900, and continued until the close of the sheep's sexual season in the 
 
 VOL. CXCVI.— B 216. 10.7.03. 
 
 Digitized by Microsoft® 
 
50 MR. F. H. A. MARSHALL ON THE CESTROUS CYCLE AND THE 
 
 within the species, whereas the period of gestation does not proportionately vary, 
 Dr. Beard's hypothesis (1897) that the time of birth is regulated by the approaching 
 termination of an ovulation interval, can scarcely be maintained.* In this connection 
 it is interesting to note that Mr. Heape (1900) finds that the length of the dioestrous 
 cycle in the rabbit may vary by as much as ten or eleven days, t 
 
 I have shown above from my own observations that the Scotch Black-faced sheep 
 when kept in the Lowlands of Scotland may have at least three recurrent periods of 
 oestrus in the absence of the ram, and flockmasters inform me that under very 
 favourable conditions there may be five or six. Mr. Heape (1900), on the authority 
 of Mr. Cameron states that Scotch Black-faced sheep have only two recurrent 
 periods of oestrus in each sexual season.| Mr. Heape further informs me that 
 Mr. Cameron's observations refer to hill sheep in the Highlands, and that the 
 sexual season for these animals is, collectively, six weeks; that is, it consists of 
 two recurrent dioestrous cycles, each of three weeks duration. Mr. Heape also 
 informs me that Mr. Cameron states that the period of oestrus with the Highland 
 Black-faced ewes occupies five days. 
 
 It can hardly be doubted that of the two conditions of the Scotch Black-faced 
 ewes that of the Highland sheep is the more natural one,§ for the sheep is essentially 
 a mountain animal in its wild state, being almost entirely confined to mountain districts 
 in the Holarctic region, and only just getting into the more tropical Oriental region. 
 " The immense mountain ranges of Central Asia, the Pamir, and Thian-Shan of Tur- 
 kestan may be looked upon as the centre of their habitat " (Flower and Lydekker, 
 1891). 
 
 What little is known about the breeding habits of wild sheep points to the 
 conclusion that at least some species are monoestrous, or at most dioestrous. The 
 accounts given by Lydekker (1898) of the rutting of Ovis musimon. 0. vignei, 
 O. ammon, and 0. canadensis show that these sheep in their wild state have a definite 
 annual sexual season, which from its short duration renders it probable that they are 
 monoestrous. Thus the sexual season in Ovis canadensis (var. typica) is described as 
 
 * According to Beard, just under ten ovulation intervals go to the sheep's gestation. 
 
 t An experienced breeder of Angus and Shorthorn cattle in Banffshire, in answer to a query regarding 
 the length of the dioestrous cycle in cows of those breeds, states that this interval varies frora two to 
 three weeks. 
 
 I Scott, writing on " The Practice of Sheep Farming," states that with hill sheep generally, the 
 " tupping season " {i.e., the sexual season) begins about November 22, and terminates with the close of 
 the year, and that the lambing season begins about April 17 and ends about May 26. These facts 
 render it improbable that there are usually more than two dioestrous cycles with hill sheep, but it must 
 be remembered that, if there are a sufficient number of rams to a flock of a given size, the ewes usually 
 succeed in getting served at an early oestrus. 
 
 § Professor EwART informs me that the duration of the' sexual season (or the number of dioestrous 
 cycles in a season) with horses varies according to the character of the breed. Thus, in semi-wild ponies 
 {e.g., Iceland or Exmoor ponies) the sexual season terminates with the approach of autumn. In the more 
 domestic types of horses, oestrus may continue to recur for a far longer period. 
 
 Digitized by Microsoft® 
 
FORMATION OF THE CORPUS LUTEUM IN THE SHEEP. 51 
 
 being about the second week of November. Similarly it may be inferred from 
 Prjevalsky's accounts (1876) that Ouis poll, 0. burrhel, and 0. argali are 
 moncestrous, and breed once a year. 
 
 Blanfoed's notes (1891) on the rutting habits of the Indian species point in the 
 same direction, for it must be remembered that the accounts given of the rutting 
 periods invariably apply to the sheep collectively in particular areas, and that it is 
 improbable that all the ewes in these areas come in season at precisely the same time. 
 Blanfoed's descriptions show further that the rutting periods of some of the wild 
 species, like those of many domestic breeds, may vary with the locality, for while with 
 Ovis vignei in the Panjdb the sexual season is September, with the same species in 
 Astor it must be very considerably later, since the young in the latter area are 
 produced about the beginning of June. (See also Lydekkee, 1898.) 
 
 The Barbary wild sheep (Ovis tragelaphus) in the Zoological Society's Gardens, 
 Regent's Park, is said to be moncestrous, breeding once annually (Heape, 1900). 
 The same is stated to be the case with Ovis burrhel in the Gardens, while the 
 Mouflon (Ovis musimon) in captivity may experience two or more recurrent 
 oestrous periods during its annual sexual season.* 
 
 Scotch Highland sheep, as already stated, are dioestrous, whereas sheep of the same 
 breed in the Lowlands resemble the great majority of the domestic breeds in being 
 polyoestrous with a single sexual season annually. Of other British breeds, Dorset 
 Horns possess the characteristic of being able to produce two crops of lambs within a 
 year, but this is discouraged by the sheep breeders, as it tends to deteriorate the 
 ewes. Hampshire Down sheep also are stated to be sometimes able to breed twice 
 annually. 
 
 The same is not unfrequently the case with ewes belonging to the Limestone breed 
 of some districts of Westmoreland and Derbyshire. This breed, although it has 
 never come into prominence, has existed for many years as an important local variety, 
 and is described by Professor Wallace (1893). Mr. Eowland Paekee, of Moss 
 End, Malathorpe, Westmoreland, from whom Professor Wallace obtained informa- 
 tion, has been good enough to inform me further regarding the sexual and lambing 
 seasons of Limestone sheep. The general lambing season is from the middle ot 
 February to the middle of March, but lambs are frequently born considerably earlier. 
 Mr. Paekee says that he has known the ewes take the ram very early when suckling 
 their lambs, a second crop being born in August, but when this occurred the ewes did 
 not breed the year after. Such an increase in the duration of the sexual season 
 is the more remarkable as occurring in north-country sheep classed as belonging to a 
 
 * These statements are from seemingly reliable information given to me by keepers in the Zoological 
 Society's Gardens, London. Subsequently Mr. F. E. Beddard, F.R.S., Prosector to the Zoological 
 Society, has very kindly obtained further information, confirming the statement that the Barbary and 
 Burrhel sheep in the Gardens are moncestrous. The Barbary sheep are described as rutting about 
 October, and the Burrhel sheep in January or February, 
 
 H 2 
 
 Digitized by Microsoft® 
 
52 MR. F. H. A. MARSHALL ON THE CESTROUS CYCLE AND THE 
 
 mountain breed, and which, although thriving well and attaining their maximum size 
 on dry lowland pasture, are hardy and active, living naturally on dry heaths or bare 
 mountain pastures (Wallace, 1893). 
 
 Although most foreign sheep resemble the majority of British breeds in showmg 
 varying degrees of polyoestrum, but a single sexual season annually, Aristotle long 
 ago recorded the fact that " in some places where the weather is warm and fine, and 
 food is abundant," sheep may have young twice a year. When this is the case the 
 spring sexual season should probably be regarded as of the nature of a continuation 
 of a series of dioestrous cycles which started at the previous autumn sexual season, 
 but were interrupted by gestation. A breeder of Dorset Horn sheep in the Isle of 
 Wight informs me that in the absence of the ram oestrus may keep on recurring at 
 regular intervals in the ewes from the autumn sexual season (during which they are 
 normally allowed to become pregnant) onwards until the following spring. It would 
 appear, however, that in these sheep there is a short anoestrous period (in the absence 
 of gestation) during part of the summer, but that its length is a variable quantity, 
 depending upon whether or not there has been a winter gestation, and being further 
 complicated by such influences as lactation, food supply, and climatic conditions. 
 
 Just as Dorset Horn ewes will come in season again within a short period ol 
 lambing, so sheep belonging to other breeds will often experience oestrus after an 
 early abortion. I have had two cases of Scotch Black-faced ewes experiencing 
 oestrus early in February. This is a great deal later than the ordinary sexual season, 
 which for this breed usually begins late in October or early in November, Unfor- 
 tunately the sheep in question had not previously been under proper observation, but 
 I think it not unlikely that they had become pregnant to a ram at an early oestrus 
 (before they came into my possession), and subsequently aborted. '■•■' 
 
 Such cases as these further justify the view that where sheep can have lambs twice 
 a year, the oestri of the additional sexual season are comparable to the oestri 
 occurring after early abortions in those sheep which can produce lambs only once 
 annually, the full series of dioestrous cycles characteristic of the respective breeds 
 being interrupted by the whole or part of a gestation. 
 
 Mr. Nelson Annandale tells me, from information given him by the Rajah op 
 Jaloe, that the horned sheep which run half wild in the old kingdom of Patani, in 
 the Malay Peninsula, normally have lambs twice every year. Mr. Annandale 
 himself saw lambs apparently a few weeks old in April and also in December. 
 
 The Merino sheep, which is almost cosmopolitan, although split up into many suId- 
 
 * Mr. Irving Dent, of Ravensneuk, Penycuik, from whom I obtained many of my sheep, informs me 
 of a case of a Scotch Black-faced ewe in his possession, which was in season in November and again in 
 the following April, giving birth to a lamb in September. The animal had become pregnant at the 
 normal time, and aborted the embryo. The case is interesting, showing that a sheep belonging to a 
 mountain breed in Scotland can, very exceptionally, come " in use " and ovulate many months after the 
 termination of the regular sexual season. The same ewe again came " in use " at the commencement of 
 the next sexual season, or at the end of the following October or beginning of November. 
 
 Digitized by IVIicrosoft® 
 
FORMATION 0¥ THE (JOKFUS LUTEUM IN THE SHEEP. 53 
 
 breeds in different places, shows considerable variation in sexual activity in different 
 parts of the world. Bonnet (1884) describes the "brunst" as returning seven 
 months after parturition; but he made no observations himself on the "brunst" 
 periods, which, as was the case with my sheep, are said to last for a short period, and 
 sometimes for only a few hours. 
 
 Among the Merino sheep of Cape Colony the general lambing season is September 
 (the spring month, corresponding to March in this country). But lambs are 
 frequently born earlier or during winter. At high altitudes, however, where the 
 ewes subsist entirely upon the natural produce of the veldt, the lambing season is 
 stated to be October, from which it may be inferred that the sexual season also 
 begins a month later than the usual time over Cape Colony. On the other hand, in 
 the low country, below the second range of mountains, there are two seasons for 
 lambing, the autumn season being in April and May. But here the lambing periods 
 are limited, since the ewes do not usually come in season until about the middle of 
 December (Wallace, 1896). 
 
 Probably the maximum amount of sexual activity reached by any sheep is that 
 attained to by the Merinos of some parts of Australia, which are described as being 
 able to breed all the year round, a fact which implies, in the absence of gestation, an 
 almost continuous series of dioestrous cycles. The report of the Chief Inspector of 
 Stock for New South Wales divides the time of lambing into six periods, which 
 embrace the whole year (Wallace, 1891). 
 
 There is, then, a complete gradation from the almost certainly monoestrous 
 condition of at least some wild species of sheep, in which there is a single sexual 
 season annually, down to the extreme of polyoestrum reached by the Merinos of 
 New South Wales, which are said to be able to breed throughout the whole year. 
 The instances cited above have been chosen mainly to illustrate this gradation, to 
 which I have not seen attention called hitherto.* 
 
 That the variability is dependent largely upon climate and food supply can hardly 
 be doubted, and the Black-faced sheep and the Merinos in different parts of Scotland 
 and Cape Colony respectively furnish direct evidence that this is the case. On the 
 other hand, there can be no doubt that the varying degrees of breeding activity are 
 also in part racial characteristics, as is shown, for example, by the Dorset Horn sheep 
 of the south of England, and more evidently by the Limestone sheep of Westmoreland 
 and Derbyshire, But that an increase in the duration (or more frequent recurrence) 
 of the sexual season is not necessarily a highly artificial condition or the part result 
 of special attention in regard to food supply, &c., on the part of the flockmaster, is 
 shown by such a condition occurring normally among the half- wild sheep of Patani. 
 
 * From Heape's notes (1900) on the subject it would seem possible that goats and pigs could be made 
 to illustrate a similar gradation from the moncestrous habit in the natural state to varying degrees of 
 polyoestrum, under domestication or in captivity, but the facts about these animals are less perfectly 
 known. 
 
 Digitized by IVIicrosoft® 
 
54 MR. F. H. A. MAESHALL ON THE (ESTEOUS CYCLE AND THE 
 
 The Scotch Black-faced sheep, further afford evidence that where the number of 
 recurrent oestrous periods is increased (beyond the normal number in the home of the 
 breed in question, e.g., the Highlands, in the case of the Black-faced sheep) the 
 duration of the oestrus is shorter,* while, whether partly or entirely as a consequence 
 of this latter fact, the length of the dioestrous cycle is also less.f 
 
 Superficial Phenomena of the Prooestrum and CEstrus. 
 
 The external signs of the prooestrum in sheep are comparatively slight. The 
 vulva is usually somewhat congested, being slightly swollen, or reddened. There 
 is, too, a slight flow of mucus from the vaginal opening, but only in one instance 
 have I observed any flow of blood. Bleeding of the uterine wall, I find to be 
 extremely slight ; but it is in the majority of cases, as I shall subsequently show, 
 present to some degree. External bleeding during prooestrum takes place with many 
 of the domestic animals, and its occasional occurrence in the sheep has been noted 
 by Wiltshire (1883) and other authors. The flow of mucus from the vaginal 
 opening, although primarily a characteristic of prooestrum, may continue during 
 oestrus, or even for a very short time during metoestrum.J 
 
 This is due to the extreme shortness of the processes of prooestrum and oestrus in 
 many domestic sheep, the mucus not having had time to escape, although the 
 
 * The longer duration of the oestrus periods of Scotch Black-faced sheep under more extreme conditions 
 (e.g., of climate and altitude) may be taken as evidence that in the perfectly wild state on the high 
 mountain, sheep have a still more prolonged oestrus which does not recur. 
 
 t A sheep breeder tells me that in Dorset Horn ewes in the south of England the cestrus may recur as 
 often as every eleven days, each oestrus lasting for only about two hours. 
 
 X That oestrus does not generally commence with bitches until external bleeding is over is a fact well 
 known to dog breeders. But a slight flow of mucus, accompanied by a little blood and a swollen vulva, 
 may continue for some days during cestrus, or even after. Such, at least, was the case with a Dandie 
 Dinmont terrier, in the possession of the present writer, for nearly a week after the performance of 
 artificial insemination. This is owing to the process of prooestrum in the bitch, although longer, being of 
 far greater severity than in the sheep, the resulting discharge consequently requiring a longer period in 
 order to completely escape. The bitch in question was ready to take the dog on the day preceding 
 artificial insemination, and gave birth to pups fifty-nine days afterwards, or sixty days after the 
 commencement of cestrus. 
 
 A breeder of Angus and Shorthorn cows (see above, note t on p. 50) informs me that with these 
 animals he has noticed a mucous discharge from the vagina hefcyre the cows come " on heat" {i.e., before 
 they are ready for service), and that heifers before coming " on heat " for the first time frequently pass 
 blood. Professor Ewart tells me that he has noticed a ^ro-oestrous mucous discharge also in the small 
 Shetland ox, which so closely resembles the "Celtic" ox [Bos longifrons). These observations are 
 of importance, as they show that the changes characterising proosstrum occur before those belonging 
 properly to oestrus, in cows as in bitches. As pointed out by Heape (1900), the prevalent confusion 
 between the two series of changes [i.e., those of prooestrum and oestrus — which are together referred to as 
 the " heat " period) in the lower Mammalia has led to false conclusions regarding the homology between 
 the prooestrum and " menstruation." 
 
 Digitized by Microsoft® 
 
FORMATION OF THE CORPUS LUTEUM IN THE SHEEP. 55 
 
 internal changes characterising these periods may be over at any rate in most parts 
 of the uterus. 
 
 The most obvious superficial phenomena presented by the internal generative 
 organs during procBstrum are the somewhat increased thickness of the uterine wall, 
 and the consequent partial obliteration of the uterine cavity, and the congestion of 
 the Fallopian tubes, which, although often somewhat coloured at other times, may 
 at the " heat " period, throughout almost their entire length, present a deep purple 
 appearance. 
 
 The presence of greatly protruding Graafian follicles from one or both of the 
 ovaries cannot properly be regarded as a characteristic of the prooestrum. 
 
 The only external indication of oestrus is that afforded by the behaviour of the 
 ewes. At this time they follow the ram, and show a certain restlessness generally. 
 At other times a ewe will refuse service. 
 
 Methods. 
 
 Immediately after killing the animal, the body cavity was opened up, and, in the 
 majority of cases, the whole generative tract, including the ovaries, was placed in a 
 10 per cent, solution of formaline, and kept there until required. The uterus was 
 usually slit open in places, in order to secure a proper penetration. Portions of the 
 uterine wall, after the uterus had been in formaline for at least six days, were 
 cut away, and, after being washed in water for about 12 hours, were hardened 
 in alcohols of gradually increasing strength. After dehydration in absolute alcohol, 
 and clearing in xylol or cedar-wood oil, they were placed in melted parafl&n (of a 
 melting point of 58° C.) for periods of time varying according to the size of the 
 specimen. The pieces of the uterine wall were then embedded in paraffin usually 
 of the same melting point, sections were cut by a Cambridge microtome at a 
 thickness varying from 6/u. to lOju,. They were sometimes attached to the slide by 
 Meyer's albumen, but generally I found the sections to become fixed to the slide 
 perfectly well after floating them on to the slide on hot water, and allowing the water 
 to dry off the slide slowly by means of a moderate heat, in the latter case without 
 employing a fixative. They were cleared in turpentine and xylol, stained, 
 dehydrated, and mounted in balsam. The stains employed were heematoxylin and 
 iron alum, hsematoxylin and eosin, methyl blue and eosin, aniline blue, and borax 
 carmine. Of these, hsematoxylin and eosin were the combination most frequently used. 
 
 The sections through the uterine wall were cut transversely to the axes of the 
 uterine cornua. Transverse sections of the Fallopian tubes were also cut and treated 
 similarly. 
 
 The ovaries were generally treated in the same way, but sometimes they were 
 cut away separately after opening the body cavity, and fixed in a saturated solution 
 of corrosive sublimate. 
 
 Digitized by Microsoft® 
 
56 ME. F. H. A. MARSHALL ON THE (ESTROUS CYCLE AND THE 
 
 Tlie Histology of the Uterus during the Dicestrous Cycle. 
 
 For general descriptions of the cotyledonary uterus, the reader is referred to 
 the veterinary text-books, to Sir William Turner's ' Lectures on the Comparative 
 Anatomy of the Placenta ' (1876), and to the various papers that have been written 
 on the cotyledonary placenta. For a short account of the transverse section through 
 the uterine wall of the sheep, Kazzander's paper (1890) may be consulted. The 
 latter author shows that the muscle layers can be essentially resolved into the 
 three usual layers, viz., an inner circular layer, a middle longitudinal layer, and 
 an outer layer containing both circularly and longitudinally arranged muscle 
 fibres. 
 
 It is also noted by Kazzander that the first muscle layer and the connective 
 tissue of the mucosa {i.e., the stroma), are not sharply divided from each other, for 
 some muscle fibres occur in the stroma. These I have noted especially in the 
 intercotyledonary stroma. 
 
 The muscle layers do not further concern the subject of this paper, except in 
 regard to the blood-vessels contained in them. These during the prooestrum are 
 liable to some amount of congestion, causing slight distension. I have never seen 
 any evidence of a breaking down of vessels in the muscle layers. 
 
 The epithelium bounding the uterine cavity consists usually of a double layer of 
 cubical cells, both on the surface of the cotyledons and of the intercotyledonary 
 mucosa. The epithelium of the glands is frequently composed of more than two 
 rows of cells, especially at their openings, where it may be many layers deep. These 
 openings do not occur on the surface of the cotyledons so much as between them, 
 and at their bases, where they are numerous. No clear line of demarcation could 
 be made out between the cells of the epithelium and those of the stroma tissue. 
 
 The stroma or sub-epithelial mucosa is a true connective tissue, and contains 
 numerous blood-vessels, glands in some cases, pigment (situated usually near he 
 surface epithelium), and deeper down in the tissue a few muscle-fibres (as already 
 mentioned), the whole being bound together in a network. The connective 
 tissue framework is, however, very much less obvious in some places (as in some of 
 the smaller cotyledons) than in others. The nuclei are considerably smaller than 
 those of the epithelial cells. I could never satisfactorily make out cell boundaries 
 in the stroma cytoplasm. 
 
 The remainder of this section of the paper is devoted to a description of the histological 
 changes through which the uterine stroma passes during a dicestrous cycle, and to a 
 brief comparison of these changes in the sheep with similar changes in other types. 
 The changes in the sheep at any rate relate primarily to the blood-vessels, They may 
 be conveniently described under four heads, as follows ; — . 
 
 Digitized by Microsoft® 
 
FORMATION OF THE CORPUS LUTEUM IN THE SHEEP. 57 
 
 (1.) Period of rest. 
 
 (2.) Period ofi;r(nvtli of stroma and increase of vessels. 
 
 (3.) Period of breakiiin' down of \essels and extiavasatioii of the blood in the 
 
 stroma. 
 (4.) Period of recuperation and pigment formatidii. 
 
 Period (I) corresponds to dioestrum, or if the diuistrous cycle in consideration be 
 the first of the sexual season, to anoestnnn. Periods ("2) and (3) both properly belong 
 to procestrum, but certain of the characteristics of period (3) not infrequently occur 
 contemporaneously with testrus. Period (4) corresponds approximately to metoestrum, 
 but recuperation and pigment formation begin actually during oestrus.* The 
 relation of these periods to dioestrum, prooistnuii, oestrus, and metoestrum depends 
 really upon the intensity or abbreviation, as the case may be, of the changes charac- 
 terising them. These changes appear to difier somewhat in intensity according to 
 the individual, while I have some slight evidence for the assertion that the severity 
 of the process tends to decrease with each successive cycle in the sexual season. 
 
 It may be at once noted that the method here adopted of classifying the sheep's 
 uterine stages is similar to that employed by Milnes Marshall (1893), and by 
 Mr. Walter Heape (1894, 1897) in grouping the menstrual periods of the human 
 species and of monkeys. The further subdivision in time of these periods in the 
 case of the sheep is scarcely warrantable, owing to the abbreviation of the prooestrous 
 process and the fact that certain characters essential to the occurrence of one period 
 may co-exist with others belonging properly to the next period, present elsewhere in 
 the same uterus. 
 
 (1.) Period of Rest, — The histological characters of the uterine mucosa at this 
 period, as compared with those of the succeeding periods, may be said to be almost 
 purely negative. 
 
 Protoplasmic processes can be seen to be passing from many of the stroma nuclei. 
 But these, though denser in some places than others, show little evidence of division. 
 The slightly denser patches appear arranged quite irregularly. 
 
 The blood-vessels are small and comparatively scarce. Those that are present 
 contain numerous red corpuscles, but I have failed to detect leucocytes. 
 
 Dark brown pigment, apparently in an amorphous state, lying free in the stroma, 
 for the most part immediately beneath the epithelium, may be present in considerable 
 quantities. According to my experience, this pigment is most frequent rather 
 between and around, the bases of the cotyledons than on their surface ; but this is by 
 no means invariable, as sometimes the cotyledons themselves, and especially the 
 smaller cotyledons, may appear superficially to be perfectly black wdth pigment. 
 {Cf. Bonnet, 1880.) 
 
 * As I shall show later, in the case of the ferret, the cestrus period may extend far into the uterine 
 recuperation stage. (See p. 61.) 
 
 VOL. CXCVI. — B. J 
 
 Digitized by Microsoft® 
 
58 MR: i\ H. A. MARSHALL ON THE (ESTEOUS CTOLE AND THE 
 
 I have never observed pigment in the uterus of a yearhng sheep, nor in any sheep 
 during the anoestrous period, the above description applying to the uterus during the 
 dicBstrous interval. Also I have found pigment to occur more frequently and in 
 greater abundance later in the sexual season (i.e., after several dioestrous cycles have 
 been passed through) than nearer its commencement ; but this statement does not 
 apply invariably. 
 
 The mode of formation of pigment is discussed under the head of Period (4), where 
 I have referred to the work of Bonnet and Kazzander. 
 
 (2.) Period of Growth of Stroma and Increase of Vessels. — The earliest indica- 
 tion of the approach of this period is afforded by a slight multiplication in the 
 number of nuclei in the stroma closest to the epithelium of the cotyledons. But this 
 change is very rapidly succeeded, if not actually accompanied, by an increase, at first 
 scarcely perceptible, in many parts of the uterus, in the size of the blood-vessels in 
 the muscle layers and the jjart of the stroma in proximity to the muscle layers. 
 
 The nuclei in process of multiplication apparently divide into two simply, and 
 without any mitoses, but of this point I could never be certain. 
 
 A little later the stroma nuclei become distributed rather more thickly, but more 
 equally, and the vessels not only are larger in size, but increased in number in nearly 
 every part of the stroma. These changes result in the uterine cavity, never large in 
 a non-pregnant sheep, becoming at this period even still smaller. 
 
 The glands tend to become somewhat swollen, while a viscous secretion may 
 frequently be detected in their slightly wider lumina. 
 
 The uterine epithelium, so far as I have observed, remains completely unaltered in 
 character. 
 
 The whole series of changes characterising this period in the case of the sheep are 
 by no means verj^ conspicuous, excepting for the continued increase in the growth of 
 the blood-vessels. This increase is shown in fig. 2 (Plate 7), where the thicker 
 distribution of stroma nuclei in the region of the epithelium of a cotyledon is also 
 represented. Superficially, the above-mentioned changes are not necessarily appai-ent 
 at all, especially in those cases where the growth is not such as to conspicuously 
 reduce the uterine cavity. 
 
 I am unable to form any opinion in regard to whether or not the increase of the 
 blood-vessels is in any sense a result of the growth of the stroma, since the two 
 processes occur almost, if not quite, simultaneously. 
 
 (3.) Period of Breaking Doion of Vessels and Extravasation of Blood in the 
 Stroma.— T\iQ congestion is followed (in most cases, at any rate) by the breaking 
 down of some of the blood-vessels, while the growth of the stroma ceases. But the 
 stroma nuclei are still in some places more closely packed than in others, and this 
 thicker distribution continues to occur most frequently in that part of the stroma 
 nearest to the epithelium hning the uterine cavity. Very frequently, the first 
 extravasation takes place from vessels situated immediately Ijelow that part of the 
 
 Digitized by Microsoft® 
 
FORMATION OF THE CORPUS LUTEUM IN THE SHEEP. 59 
 
 stroma where the nuclei are thickest. The vessels in the muscle layers and the 
 adjoining part of the stroma do not rupture at all, so far as I have observed, and the 
 same can be said for the congested vessels of the Fallopian tubes. The rupture in the 
 more superficial portion of the stroma is apparently the direct result of congestion 
 and the consequent strain on the vessel walls. 
 
 The blood corpuscles thus set free become scattered in the stroma, where they form 
 irregularly shaped patches and sti-eaks lying a little below the epithelium. Leucocytes 
 are also extravasated with the red corpuscles, but the proportion of the former to the 
 latter, so far as I am able to judge, is not abnormal. I lia\'e no evidence of the 
 existence of wandering cells in the uterine stroma apart from what might quite well 
 have been expelled during the rupture of the blood-vessels. 
 
 In fig. 3 (Plate 7) is shown part of a section of the mucosa of a cotyledon, with 
 blood-vessels congested, but not ruptured, somewhat deeper in the stroma, and 
 extravasated blood, resulting apparently from very recently broken down vessels 
 situated nearer the epithelium. 
 
 Fig. 4 (Plate 7) represents a section showing the nearest approach to a blood 
 lacuna that I have observed in a sheep's uterus during prooestrum. The disappear- 
 ance of the epithelium at this point is, I think, in part due to faulty manipulation, 
 since it is unusual to find more than about half-a-dozen cells wanting at points where 
 small streams of blood corpuscles were being poured into the uterine cavity. The 
 mere fact, however, of the disappearance of the epithelium opposite a space that was 
 obviously a miniature lacuna, whether it had become denuded naturally or otherwise, 
 is evidence of a weakness at this jDoint in the epithelial wall, due to the pressure 
 exerted from Avithin by the extravasated blood. It is noticeable that in this section, 
 although the stroma presents a fairly normal appearance, capillaries are still wholly 
 absent, while a few extravasated corpuscles are seen scattered in the superficial 
 portion of the tissue. 
 
 Formerly I was of opinion that denudation (if only to a slight extent) of epithelium 
 and bleeding into the uterine cavity occurred invariably during the sheep's prooestrum. 
 But material subsequently obtained points to the conclusion that this is not neces- 
 sarily the case, although it is by no means easy to distinguish between a uterus in 
 which the prooestrum is nearly over and in which there has been little or no bleeding 
 to the exterior, and another uterus in which rupture of vessels has only just begun. 
 But where cestrus has commenced and the uterine epithelium shows every evidence 
 of having remained intact during the prooestrous process, and the blood-vessels deeper 
 in the stroma are not much congested, while blood in an early stage of haemorrhage 
 is seen in the superficial tissue, it is safe to assume that the severity of the process 
 was not sufiiciently great to induce bleeding into the cavity. 
 
 According to my observations, the severity of the prooestrous process tends to 
 diminish with each successive dioestrous cycle in the rutting season. Sometimes in a 
 late prooestrum it would appear probable that Period (3) is not reached at all, but 
 
 I 2 
 
 Digitized by IVIicrosoft® 
 
60 MR. F. H. A. MAKSHALL ON THE (ESTEOUS CYCLE AND THE 
 
 that the congested vessels subside without any rupture. But the matter is compli- 
 cated by the undoubted fact that in one part of the uterus a more advanced stage 
 may be reached than in another part of the same uterus. Bleeding into the uterine 
 cavity, so far as I have seen, is more frequent through the cotyledons than between 
 them, and is commoner in the case of large cotyledons than with smaller ones. 
 
 I have never seen any denudation of stroma during the sheep's prooestrum. The 
 lumina of the glands during this period tend to become still wider, while the glandular 
 secretion may increase in quantity. 
 
 (4.) Period of Recuperation and Pigment Fo7'mation. — The characteristics of this 
 period (as indicated in sections through the mucosa) depend upon the severity of the 
 changes that have occurred in the preceding period, not only in the uterus as a whole, 
 but in the particular j^art of it under consideration. Where bleeding into the uterine 
 cavity has been considerable, there may be little or no extravasated blood retained in 
 the stroma, and, since bleeding is commoner (at least according to my experience) in 
 the upper parts of the cotyledons than near their bases and between them, one might 
 expect to find less extravasated blood retained in the former parts of the stroma than 
 in the latter. This is by no means invariable, but, as a general rule, I have found it 
 to be the case. 
 
 The new epithelium at places where in the preceding period bleeding into the cavity 
 was going on, takes its origin from the adjoining epithelium which did not become 
 denuded. In this way the re-formation of the epithelium in the case of the sheep is 
 easily accounted for, since, as already remarked, only a very inconsiderable number of 
 epithelial cells disappear during the sheep's prooestrum. 
 
 Fine branching protoplasmic processes, arising apparently from the nuclei of the 
 stroma, are commonly seen in the early part of the recuperation period, but not where 
 the breaking- down process has been less severe. These processes disappear, to a great 
 extent, at a later stage. The mucosa, as a whole, undergoes a slight shrinkage during 
 the early part of Period (4), but this begins really at the time when bleeding into the 
 cavity occurs. Congestion gradually becomes less and less in the deeper parts of the 
 uterine wall and in the walls of the Fallopian tubes, the vessels being at length 
 reduced to their normal proportions. Meanwhile nuclear division in the stroma is of 
 rare occurrence. Blood-vessels in the superficial part are rarely found in the earlier 
 recuperation stage, and there is no evidence that any of the extravasated blood 
 corpuscles which remain in the stroma are gathered up afresh into the circulatory 
 system, as Mr. Heape (1894) states to be the case after menstruation in monkeys. 
 But at a period three days after cestrus it would appear that new capillaries have 
 been formed. 
 
 The extravasated blood corimscles retained in the stroma foim j.igment, as supposed 
 by Bonnet (1882), although at the time of writing the paper referred to Bonnet 
 does not appear to_have observed extravasated blood in the sheep's uterus. He 
 supposed the extravasation to take place in the deeper part of the tissue during the 
 
 Digitized by IVIicrosoft® 
 
FORMATION OF THE CORPUS T.UTEUM IN THE SHEEP. (i 1 
 
 '* brunst," the derivatives of the corpuscles being carried in the form of pigment to 
 the more superficial portion of the mucosa by wandering cells. Subsequently he 
 describes these pigment-bearing lencocytes, wliich he regards as true pigment cells, as 
 breaking down, leaving the pigment free in the mucosa, in the form of light-brown, 
 coarse-grained masses of colouring matter. Kazzander (1890) seems to have been the 
 first to have seen the extravasated blood which he correctly describes as occurring 
 mostly in the superficial tissue. But Kazzander does not admit the agency of 
 leucocytes, while the appearance of pigment cells he describes as being due to the 
 arrangement of pigment granules around certain of the stroma niiclei. He regards 
 the process of transformation of extravasated blood corpuscles into pigment as 
 occurrinp; free in the tissue and outside the cells. 
 
 All gradations from freshlv-extravasated blood to completely transformed pigment 
 appear in mv sections, but I cannot adopt Kazzander's view that the change takes 
 place independently of wandering cells. On the other hand, the crowding togetlier 
 of little groups of extravasated corpuscles in the early stages often plainly associated 
 with a leucocyte, and the subsequent arrangement of pigment granules in cells having 
 all the appearance of pigment cells, point to the opposite conclusion. At a later 
 period the pigment generally becomes so thickly distributed both immediately below 
 the uterine epithelium and elsewhere, that it is impossible to state definitely Avhether 
 it is within cells or not, but it has the appearance of lying free in the stroma. 
 
 Though the evidence is not quite clear, I find it difficult to resist the conclusion that 
 the transformation from extravasated corpuscles into pigment is worked through the 
 agency of leucocytes which, when laden with pigment, appear as pigment cells. 
 Leucocytes, undoubtedly, are extravasated with the red corpuscles, though not, in the 
 case of the sheep, in any considerable number. In the menstruation of monkeys 
 they are extravasated in very large numbers, and, apparently, quite unnecessarily 
 (Heape, 1894). It would appear as if in the latter case the need for leucocytes is 
 obviated by the red corpuscles which remain in the stroma (after denudation), together 
 with the leucocytes being gathered u]? anew into the circulatory system (Heape, 1894). 
 So far as I am aware, there is no record of pigment formation in the uterine mucosa 
 of monkeys. 
 
 The amount of pigment formed in the sheep's sexual season depends upon two 
 factors ; firstly, upon the extent to which extravasation has occurred in the prooestrous 
 periods,* and, secondly, upon the quantity of blood retained in the stroma (or, con- 
 versely, upon the extent or existence of bleeding into the uterine cavity). Fig. 6 
 (Plate 7) rejoresents part of a section through the mucosa of a sheep killed towards 
 the end of the sexual season, i.e., a sheep that had passed through several recurrent 
 dioestrous cycles. The animal was killed four days after oestrus. The large quantity 
 of pigment formed was no doubt derived from blood which had been extravasated 
 
 * The quantity of pigment formed must also, of course, depend upon the number of prooestrous periods. 
 This may be inferred from the facts stated below in the same paragraph. 
 
 Digitized by IVIicrosoft® 
 
'i2 MR. F. H. A. MARSHALL ON THE (ESTROUS CYCLE AND THE 
 
 during the several prooestrous periods, and not merely during the most recent of the 
 series. Derivatives of hlood corpuscles, associated with wandering cells, by which 
 they were apparently in process of being taken up, and dating from the last 
 prorostrum, AA'ere still recognisable. Elsewhere in the same section hciematoidin crystals 
 were found to occui-. 
 
 Literature of the Proce.^trmn. — The early papers Ity Bonnet (1880, 1882), and 
 Kazzander's paper (1890) on pigmentation and its mode of formation in the sheep's 
 uterus have been already referred to. Subsequently Bonnet (1892), has noted 
 the existence of swelling, congestion, and glandular secretion during the " brunst " of 
 many domestic animals, and made the further statement that in the mare and the 
 sow and ruminants bleeding may take place in the mucosa. Ellenbergjer, in the 
 same volume (1892), makes similar statements, and although not describing extra- 
 vasation in the mucosa, refers to the pigmentation there in the case of ruminants. 
 (See also Bonnet and Kolster, 1902.)* 
 
 For an account of the evidence of the existence of the prooestrous period in the 
 Mammalia generally, and a resume of the literature, the reader is referred to 
 Mr. Heape's recent paper (1900), where the question o± the homology between the 
 procestrum and " menstruation " is also dealt with. 
 
 Mr. Heape (1894, 1897) groups the phenomena of menstruation in Semnopithecus 
 ciitellus and Macac)i,s rhesus into four periods and eight stages. For purposes of 
 comparison they are set out here : — 
 
 A. Period of rest. 
 
 Stage T. — The resting stage. 
 
 B. Period of growth. 
 
 Stage II. — The growth of the stroma. 
 Stage III. — The increase of vessels. 
 
 C. Period of degeneration. 
 
 Stage IV. — The breaking down of vessels. 
 
 Stage V. — The formation of lacuna. 
 
 Stage VI. — The rupture of lacunae. 
 
 Stage VII. — The formation of the menstrual clot. 
 
 D. Period of recuperation. 
 
 Stage VIII. — The recuperation stage. 
 
 * Since concluding my paper, 1 have seen Kolster's memoir in the last issue (Dec. 1902) of the 
 'Anatomische Hefte,' where pigment cells in the uterine mucosa of the sheep and cow at the "brunst" 
 are described, and Bonnet's original descriptions are to a great extent confirmed. Kolster o^ives a 
 very short account of the histology of the uterus of the sheep, cow, and pig at the "brunst," while 
 Bonnet (1902) in the same number briefly describes the uterus of the bitch at this period. 
 
 Digitized by IVIicrosoft® 
 
FORMATION OF THE CORPUS LUTEUM IN THE SHEEP. 63 
 
 As already remarked, the four divisions into wliich I have grouped the diojytrous 
 changes through which the shee])'.s uterus passes are identical with the periods into 
 which Mr. Heape has divided the menstrual cycle in monkeys. Tlie chief differences 
 relate to tlie comparatl\'e slightness and corresponding abhreviation of the procestrous 
 process iu the sheep. Thus the growth of the stroma in the sheep is very much less 
 than with the menstriial cycle of monkeys, and occurs simultaneously with the 
 increase of vessels ; and hi Period (3), although the characters of Stages IV., V., 
 and VI. may all he present, they do not necessarily occur in succession throughout 
 the whole uterus, for while one part of the uterus may pass through Stage VL, in 
 another part bleeding into the uterine cavity may never take place at all. The most 
 important distinction between the two processes lies in the formation of the 
 menstrual clot (Stage VII.) in the Primates. This results from the intensity of the 
 bleeding and the severity of the denudation, which in the case of the sheep's 
 procestrum scarcely occurs, being, as I have shown above, confined to certain spots 
 on the surface of the mucosa,, where a fe^^' epithelial cells may be removed. The 
 period of recuperation with Semnopithecus and Macacus, as described by Mr. Heape, 
 differs from that following the procestrum in the sheep, mainly in the gathering-up 
 into the circulatory system of the extravasated blood corpuscles and leucocytes, and 
 the consequent absence of pigment formation.'"" 
 
 That the prooestruni of the sheep and the menstrual phenomena of monkeys are 
 physiologically homologous is rendered exceedingly probable in view of the facts 
 cited above, and without reference to other members of the mammalian class. But 
 the probability seems to become converted into a certainty that can hardly be 
 doubted, after making a further comparison with the procestrous processes of the 
 bitch as described by Ketterer (1892), and of the ferret as observed by the present 
 writer. Special reference must also be made in this connection to Stratz's notes 
 (1898) on the "menstrual" discharges of Ta2)aia and Tarsius, in each of which a 
 blood-clot is described, although there appears to be no removal of mucosa tissue 
 beyond the epithelium. 
 
 The only account hitherto published of a systematic investigation into the 
 phenomena of the procestrum in a lower mammal, so far as I am aAvare, is that ol 
 Retterer (1892), who briefly describes the process in the case of the bitch, as just 
 mentioned. 
 
 The uterine mucosa, according to IIetterer, at the beginning of the procestrum 
 grows to several times its previous thickness. The blood-vessels increase in number 
 and in size, and in the superficial part become so much distended as to cause rupture, 
 which results in a considerable extravasation of blood. Lacuna-like spaces are 
 
 * In his later paper (1900) to which I have referred, Heape has extended this scheme of classification 
 to the procestrous phenomena of the lower Mammalia. In another paper (1898) he has described two 
 menstruating uteri from the human female, showing characters comparable respectively to Stages TV. and 
 VII. of the monkey. 
 
 Digitized by Microsoft® 
 
64 MR. r. H. A. MARSHALL ON THE (ESTROUS CYCLE AND THE 
 
 formed, and tlies3 discharge the contained blood into the cavity of the uterus. 
 Eettereb refers also to pigment formation from extravasated blood corpuscles in the 
 mucosa. Mr. Heape (1900) also states that he has investigated the processes of the 
 bitch's prooestrum to some extent, and that his results confirm those of Retterer. 
 From these accounts it appears probable that there is no removal of epithelium in 
 the case of the bitch excepting at points where bleeding into the cavity has 
 occurred. 
 
 The CEntron^ Cijcle in the Ferret. — I append below, for purposes of comparison, a 
 very short account, which may be considered as of the nature of an advanced abstract 
 of my observations on the oestrous cycle in the ferret. 
 
 The ferret is moncestrous, and has one annual sexual season, which, in the absence 
 of the male, consists of a single greatly prolonged cestrous cycle. The prooestrum 
 alone, during which the female will I'efuse the male, may extend for over three 
 weeks, while the oestrus, or period during which the female Avill allow copulation, 
 may last for a further period of five or six weeks, and in some cases, if pregnancy be 
 not permitted, seems to extend almost indefinitely. But in a ferret in which oestrus 
 was observed in the beginning of June (at the time when it was procured), the " heat" 
 period was completely over at the end of the first week of July, coition not having 
 been permitted. The sexual season is ordinarily in the spring and first part of the 
 summer. Five bitch ferrets which were obtained in October showed no signs of 
 coming in season until the end of the following March. 
 
 Apart from the attitude of the animals towards the male during the oestrus, the 
 only external indications of the existence of "heat" are the very marked swelling 
 of the vulva during both prooestrum and ojstrus, and the occurrence of a sanguineo- 
 mucous flow from the genital aperture ; the latter characteristic marking especially 
 the close of the prooestrous period, although it may continue to a slight extent 
 after the commencement of oestrus. With two or three individuals I failed to 
 note external bleeding, but it may have taken place Avithout being observed, as 
 it was sometimes impracticable to make regular observations upon the animals 
 during their prolonged oestrous cycle. 
 
 The oestrus associated with the swelling of the vulva extends far into the uterine 
 recuperation stage, if not sometimes beyond it ; so that similar changes to those 
 which in the sheep characterise metoestrum, in the ferret may occur contemporaneously 
 with oestrus.* 
 
 The histological charactei'S of the prooestrum Avith the ferret are essentially like 
 those described by Retterer for the bitch. There is a well-marked groAvth period, 
 during which the uterine cavity is reduced to about half its normal size, the mucosa 
 being correspondingly increased in thickness. The vessels subsequently become 
 
 * Bears and certain other animals in the Zoological Society's Gardens are said to permit copulation at 
 almost any time; Lnt this is an abnormal condition, and due to the influences of captivity (Heape, 1900). 
 Regarding the occuiTeuce of o\-ulation in the ferret, sec p. 70 in the next section of this paper. 
 
 Digitized by Microsoft® 
 
FORMATION OF THE CORPUS LUTEUM IN THE SHEEP. 65 
 
 extoemely dilated with blood. At a later stage lai'ge quantities of extra vasated blood 
 corpuscles are found scattered in the mucosa, not merely in the superficial part, but 
 deeper in the tissue. But in the ferret also there is no breaking down of vessels in 
 the muscle layers of the uterine wall. The growth and breaking down stages are 
 characterised further by a distinct swelling of the glands. These changes are followed 
 by a denudation stage which seems to be more marked than the corresponding stage 
 in the bitch. In one specimen large parts of the mucosa were found to be stripped of 
 the lining epithelium, and in this case the denudation could hardly be ascribed to 
 faulty manipulation, since numerous groups of detached epithelial cells, in a more or 
 less degenerate condition, and accompanied by a quantity of mucous secretion 
 derived from the glands, were found free in the uterine cavity. In a part of the 
 same uterus there was evidence also of a prooestrous removal of stroma tissue. I find 
 no indication, however, that the loose tissue and blood corpuscles are ever aggregated 
 into a prooestrous clot, and I am inclined to think that the discharge thus composed 
 is usually got rid of very gradually. 
 
 In the recuperation period which follows, the stroma is seen to be much reduced, 
 and at the same time to be less compact, while the cavity at the beginning of this 
 period is larger than at any other times during the cycle. This reduction of the 
 stroma tissue is by itself some evidence of removal of its most superficial portion 
 during the denudation period. The blood corpuscles which were retained in the 
 stroma eventually disappear in a manner that might be interpreted as indicating that 
 they are gathered up afresh into blood-vessels, but in the absence of a quite complete 
 series of stages I cannot be certain of this point. But it is evident that they do not 
 become transformed into pigment as with the sheep. A new (or partly new) epithelial 
 layer, composed of flattened cells, is formed during recuperation, and in an early 
 stage of this period I have observed its mode of formation from adjoining epithelial 
 cells in spots where the stroma was still bare. Whether or not at a still earlier stage 
 any portion of the epithelium is renewed from the underlying tissue, as described to 
 be the case for monkeys by Mr. He ape (1894, 1897), I am unable to say positively, 
 but the relation between certain tracts of epithelium and the adjoining stroma 
 suggests the correctness of this view in the case of the ferret. 
 
 It is clear from the facts summarised in the preceding paragraphs that all the stages 
 into which Mr. Heape first divided the menstrual cycle of Semnopithecus are repre- 
 sented in the CEstrous cycle of the bitch (Rettbrer, 1892), and in that of the ferret, 
 with the exception of Stage VII. (the formation of the menstrual clot), the characters 
 of which in the case of the latter animals are only partially recognisable. It is 
 apparent also that the prooestrous process in the carnivore, in regard to its severity, 
 is approximately intermediate between that of the sheep, as described above, and the 
 menstruation of monkeys, as worked out fully by Mr. Heape. But the oesti'ous 
 cycle of the ferret difiers both from the dioestrous cycle of the sheep and from the 
 menstrual cycle in the Primates in its far more extended duration. 
 
 VOL. CXCVI. — B. K 
 
 Digitized by IVIicrosoft® 
 
66 MR. F. H. A. MARSHALL ON THE GESTROUS CYCLE AND THE 
 
 Ovulation, ivith Notes on the Atretic Follicle, and the Causes of Barrenness. 
 
 Ovulation in the sheep, as in the majority of the lower Mammalia, so far as is known, 
 takes place during the sexual season. In the early summer months (the anoestrous 
 period with the breeds studied) the ovaries are more compact than later in the year, 
 and show little or no indication of maturing follicles or corpora lutea. With the 
 approach of autumn, the Graafian follicles are seen in varying degrees of protrusion 
 from the surface of the ovaries, while in the sexual season ripe (or nearly ripe) and 
 recently ruptured follicles can be found in the ovaries of nearly every non- 
 pregnant ewe. 
 
 The ovaries of lambs born in the previous spring, as autumn approaches present 
 similar superficial changes, the sexual season with them beginning about the same 
 time as that of older sheep. 
 
 In sections through ovaries of sheep killed during the anoestrous period it is also 
 unusual to find follicles in early stages of degeneration. Those that do occur, so far 
 as I have observed, are follicles that have only reached a comparatively early stage 
 of growth, and not such as have matured and failed to rupture. 
 
 When ovulation has recently taken place, it can usually be readily detected super- 
 ficially by the presence of a blood-clot on the surface of the ovary, or the existence of 
 a pit-like depression indicating the point of rupture of the follicle, which remains open 
 for at least four days. But to avoid all possible confusion, of which in other animals 
 there has been much, between true corpora lutea in their early stages of development 
 and atretic follicles which have never ruptured, I considered it advisable to undertake 
 further the study of the changes undergone by the discharged follicle in the sheep. 
 These changes are described under the heading, " The Formation of the Corpus 
 luteum" (p. 78). 
 
 My observations regarding the stimulus necessary to induce ovulation may be 
 conveniently described in the form of a series of experiments. As already mentioned, 
 the periods of oestrus were detected by the attitude of the ewes towards the ram, 
 which, when required for the purpose of the experiment, was rendered temporarily 
 incapable of insemination. 
 
 Experiment 1. — A virgin ewe (a cross between a Cheviot ram and a Border 
 Leicester ewe) was killed a day after that its first oestrus had been observed, 
 coition not having been permitted. It was found that ovulation had recently taken 
 place. The subsequent examination of the ovaries pointed to the fact that there had 
 been no previous ovulation. The histology of the discharged follicle is described in 
 the following section of this paper as a twenty- four hour stage in the development of 
 the corpus luteum. 
 
 Experiment 2.— A Scotch Black-faced sheep early in the breeding season was 
 noted to be in a state of oestrus, but coition was not permitted. Two days later 
 the ewe was killed, when it became evident that a follicle had recently ruptured. 
 
 Digitized by Microsoft® 
 
FOEMATION OF THE COKPUS LUTEUM IN THE SHEEP. 67 
 
 There was no evidence that ovulation had occurred previously during that sexual 
 season, neither had a previous oestrus been recorded that year. The sheep was a 
 four-year-old, and had given birth to a lamb the previous year. This experiment, 
 then, was a repetition of Experiment 1, but with a sheep that was not a virgin, the 
 result in regard to the occurrence of ovulation being the same in each instance. The 
 discharged follicle is described below as a fifty-hour stage in the formation of the 
 corpus luteum. 
 
 Experiment 3. — (Estrus was noted in a Scotch Black-faced sheep shortly after the 
 beginning of the sexual season. After an interval of fifteen days, oestrus was again 
 noted. The sheep was killed four days later, when a newly discharged follicle was 
 found in one of the ovaries. By a comparison with other stages in the formation of 
 the corpus luteum, it was stated that the foUicle in question had ruptured about the 
 time of the second oestrus observed. Coition had not been permitted during either 
 oestrus. 
 
 Experiment 4. — (Estrus was noted in a Black-faced ewe, and was observed to 
 recur after fifteen days, and to recur again after a further interval of sixteen days. 
 The sheep was killed about seven hours after that the third recurrent period of 
 oestrus had been noted, and vtdthout being served. Ovulation had very recently 
 taken place. The newly discharged follicle is described below (p. 79) as a seven-hour 
 stage in the formation of the corpus luteum. I found subsequently in the same ovary 
 what I beHeve to be the remains of a follicle that had discharged during one of the 
 previous oestrous periods. 
 
 This and the preceding experiment show that sheep can ovulate independently of 
 coition at other oestrous periods than the first of the sexual season. In neither 
 experiment were the sheep virgin animals. 
 
 Experiment 5. — (Estrus was noted in a Black-faced eWe (not a virgin) in the 
 beginning of February, and the ewe was served by a ram of its own breed. Sixteen 
 hours after coition the ewe was killed, and a very recently discharged follicle was 
 discovered. Subsequently, on cutting one of the ovaries into sections, I found (in 
 addition to the young corpus luteum which is described in the next section) a 
 degenerate foUicle, which, judging from its size and position, which were those of a 
 nearly ripe follicle, and its state of degeneration had probably attained complete 
 ripeness before undergoing atresia. The membrana granulosa had almost entirely 
 disappeared, its few remaining cells being in an advanced stage of degeneration. 
 Very little ingrowth of connective tissue from the thec« had occurred, and what had 
 was looser and more irregular than the normal ingrowth. The cavity contained a 
 fluid resembling the liquor of the Graafian follicle, but staining somewhat more deeply. 
 There was no blood-clot, such as sometimes occurs in atretic follicles in the rabbit's 
 ovaries, nor any indication that there had been a breaking down of vessels, or rupture 
 to the exterior. No recognisable trace of the ovum was found, but this may have 
 been due to the fact that some sections were unfortunately lost in the cutting. 
 
 K 2 
 
 Digitized by IVIicrosoft® 
 
68 MR. F. H. A. MAESHALL ON THE (ESTEOUS CYCLE AND THE 
 
 The special interest attaching to this observation lies in the possibility ot the 
 atretic follicle having attained complete maturation and afterwards undergone 
 degeneration owing to want of sufficient stimulus to induce ovulation. 
 
 Experiment 6.— (Estrus was noted in a Scotch Black-faced ewe (not a virgm)^ m 
 the beginning of February, but coition was not permitted to occur. On the foUowmg 
 day the ewe, which was no longer in season, was killed. There was a considerably 
 protruding follicle in one ovary, but no indication of a recently ruptured follicle in 
 either ovary. On the other hand, it was quite clear that ovulation had not occurred 
 during the observed oestrus. The animal had not been under strict observation 
 before this oestrus, so it is unknown how many dioestrous cycles had been passed 
 through. 
 
 In three other experiments in which the sheep were killed within seven hours 
 of coition, ovulation had not yet taken place, while in a further case also in which 
 the sheep was killed about twelve hours after the detection of oestrus, ovulation had 
 not at that time occurred. But in the latter experiment coition was not permitted. 
 In no cases, other than those mentioned above, after that a sheep in which oestrus 
 had been noted shortly before was killed, did I fail to find one or more recently 
 discharged follicles.'"' 
 
 It may be concluded from the above-described experiments that hill sheep, or half- 
 bred hill sheep, whether virgins or otherwise, under ordinary conditions, ovulate in 
 response to a stimulus during oestrus (or possibly occasionally during the prooestrum), 
 and that they can do so independently of coition during any oestrus of the regular 
 sexual season ; but that ewes in which, for one reason or another, oestrus occurs at a 
 considerable time after the termination of the regular sexual season, may require 
 the additional stimulus imparted by sexual intercourse in order to bring about 
 ovulation. There is also evidence that where coition occurs at the beginning of an 
 oestrus period, it may provide a stimulus inducing ovulation to take place a few hours 
 earlier than it might in the absence of coition ; or, in other words, that if ovulation 
 has not already occurred during an oestrus, the stimulus derived from the act of 
 coition may, so to speak, hasten the ovulation. 
 
 The statement concerning the non-occurrence of spontaneous ovulation in sheep 
 during oestrus at other times than the normal sexual season requires further explana- 
 tion. 
 
 On the supposition that the sheep mentioned in Experiment 6 had come " in use " 
 at the usual time for Black-faced sheep, kept under favourable conditions, and that 
 
 * The following statement occurs in the 'Preliminary Communication,' p. 137 (1901): "In no case 
 after a sheep in which oestrus had been observed was killed in order to obtain a stage in the development 
 of the corpus luteum was the corpus lutemn not found." At the time of writing this passage I had 
 obtained only one very early stage (a " seven-hour stage ") in the formation of the corpus luteum. 
 Subsequently I have had considerable diflSculty in getting very early stages, owing to their being no fixed 
 time for ovulation during oestrus. 
 
 Digitized by Microsoft® 
 
FORMATION OF THE CORPUS LUTEUM IN THE SHEEP. 69 
 
 each of the dicestrous cycles lasted sixteen days, the observed oestrus was the 
 seventh of the sexual season. This would imply a more extended series of dicestrous 
 cycles than is usual among ewes of this breed kept under any conditions. But it may 
 be that while some ewes have such an increased breeding capacity, yet the stimu- 
 lating power at their disposal duiing an exceptionally late oestrus is so reduced that 
 without coition and the presence of speimatozoa in the uterus it is insufficient to 
 cause ovulation. On the other hand, the ewe may not have come " in use " first 
 until very much later than the usual season, in which case the need for a special 
 stimulus was associated with an individual lack of breeding capacity. The only other 
 alternative accounting for the late oestrus, and one which from other evidence appears 
 to me the most probable, is that the sheep became pregnant to a ram at an early 
 oestrus, and soon after aborted, subsequently coming in season again. This explana- 
 tion, also, implies a want of generative vigour, whether temporary or otherwise, on 
 the part of the ewe in question. 
 
 Either of these explanations will account similarly for the presence of the atretic 
 follicle described in Experiment 5. This experiment also serves to show that with 
 the additional stimulus set up by sexual intercourse, ovulation can occur during an 
 oestrus outside the normal sexual season. 
 
 Whether or not the relation between copulation and ovulation is the same for all 
 breeds, I have no means of knowing. But seeing that it can vary with circumstances 
 with one breed, it is possible that different breeds show different degrees of adjust- 
 ment in this respect. However this may be, the fact of the occurrence of spontaneous 
 ovulation during oestrus in sheep, belonging to some breeds at least, is of importance, 
 since it indicates the possibility of obtaining successful results by artificial insemina- 
 tion. 
 
 According to Hausmann's description (1840) it would appear that sheep ovulate 
 normally as a result of coition, but this conclusion may be due to the fact that 
 Hausmann's observations were upon sheep which had been served before ovulating ; 
 and, as I have shown above, copulation not infrequently does occur prior to 
 ovulation. 
 
 Ovulation in other Mammals. — The rabbit, it would appear, is the only other animal 
 which is known to vary in regard to the stimulus necessary for ovulation. Mr. Heape's 
 experiments (1897) show that in the virgin rabbit both copulation and the presence 
 of spermatozoa in the uterus may be essential, while the number of atretic follicles to 
 be observed in almost any rabbit's ovaries renders it probable that it is not only 
 virgins which fail to ovulate spontaneously when " on heat." Flemming (1885) states 
 that he found at least twenty degenerating follicles in each of the ovaries of two 
 virgin rabbits, and also records atretic follicles in an animal that had given birth to 
 young, and in two old animals. On the other hand, Weil (1873) makes the state- 
 ment that ovulation can occur without coition in doe rabbits shortly after parturition, 
 arid Ott (1882), and, more recently, Gbusdew(1896) have obtained successful results 
 
 Digitized by Microsoft® 
 
70 ME. F. H. A. MAESHALL ON THE GESTEOUS CYCLE AND THE 
 
 from the artificial insemination of rabbits, but not in a high percentage of cases. 
 Professor Ewart (1901) has described an experiment with a grey doe rabbit (a virgin) 
 which was served (although not in season) forty days before parturition, or ten days 
 before ovulation, thirty days being the usual period of gestation with the rabbit. 
 This case seems to show that sometimes, even with virgin does, the presence of sper- 
 matozoa in the rabbit is enough to induce ovulation, for the ovulation can scarcely 
 have occurred as a result of a stimulus imparted by coition, since the rabbit was 
 served ten days previously, and before it had come in season.'" 
 
 Most other mammals in which the subject has been investigated are said to ovulate 
 spontaneously during oestrus. Sobotta (IS 95) has shown that this is so with mice, 
 while it follows that those animals which can become pregnant from artificial insemi- 
 nation during cBstrus, also can ovulate without coition. Such is the case with mares 
 and donkeys and cows (Heape, 1897, 1898), t with which artificial insemination is now 
 fairly widely practised. Artificial insemination has also been performed with suc- 
 cessful results on bitches (Millais, 1895 ; Heape, 1897, &c. ; see also note on p. 54), 
 and also on the human female (Heape, 1897, &c.), but not, so far as I am aware, upon 
 other Mammalia. 
 
 From my observations on ferrets, I have found that ovulation does not take place 
 without coition, and, unless the female ferret is warded at the right time for ovulation 
 to occur, the follicles and contained ova undergo atresia. But, notwithstanding the 
 ovulation being abortive, the oestrus pei'sists until long after the time for ovulation 
 has gone by, so that a ferret, when warded too late in the season, fails to become 
 pregnant. This may be the explanation of the fact noted by Robinson (1893) that 
 coition very frequently does not result in pregnancy, although the ferret may have 
 copulated more than once during oestrus. Thus the persistence of the oestrus which 
 continues far into the recuperation stage of the uterus, or beyond it, is associated 
 with numerous degenerating follicles in the ovary, which, in the natural condition, 
 would no doubt have discharged at an earlier period, under the influence of the 
 increased stimulus of coition. A female ferret which I inseminated artificially during 
 oestrus did not become pregnant, owing, probably, to the presence of the sperms in 
 the uterus without the act of coition failing to induce ovulation. 
 
 In many bats ovulation is described as occurring at quite a different time of the 
 year from oestrus, copulation generally taking place in autumn and ovulation in 
 spring (van Beneden and Julin, 1880), there being sometimes as long as six months 
 between them. The sperms are said to retain their vitality until the spring ovulation. 
 
 * The doe here referred to was a quarter-bred wild rabbit, and in colour and disposition resembled its 
 wild grandparent. It has occurred to me that this fact possibly may explain the occurrence of ovulation 
 in a virgin doe without the additional stimulus of coition. A wild rabbit may conceivably have greater 
 stimulating power at its disposal than one might expect to find evidence of in breeds that have long been 
 domesticated. 
 
 t Donkeys are not referred to by Mr. Heape in this connection, but artificial insemination has been 
 performed on donkeys to my knowledge, and has resulted in pregnancy. 
 
 Digitized by IVIicrosoft® 
 
FOEMATION OF THE CORPUS LUTEUM IN THE SHEEP. 71 
 
 It follows that in bats ovulation can occur without even the stimulus of oestrus, 
 unless the existence of the spermatozoa in the uterus may in some way provide a 
 stimulating influence. But ovulation could only occur if there were ripe follicles 
 present in the ovary, and the ovary in the winter months (during hibernation) is said 
 to be in a state of quiescence (van der Stkicht, 1901). Lately, however, Salvi 
 (1901), from observations upon the bats in the Grotta dell 'Inferno, near Sassari, says 
 that they copulate in spring as well as in autumn, the autumn sexual season having 
 long been known. 
 
 According to Stratz (1898) the ova of Tupaia become mature at " menstruation." 
 In Sorex also the mature ova are produced during the " heat " periods, while at other 
 times the follicles undergo atresia (Stratz, 1898). 
 
 The periods for the occurrence of ovulation in the Primates have been the subject 
 of a considerable amount of controversy, the question being discussed at some length 
 in three papers by Mr. Walter Heape (1894, 1897, 1898), who shows that men- 
 struation and ovulation are to a very large extent independent functions. Mr. Heape 
 points out further that whereas, in both monkeys and the human species, menstrua- 
 tion may occur periodically all the year round, in monkeys there is a limited time 
 for conception and ovulation, while in civilised woman this period is not limited to 
 any particular time of the year, although there is evidence that primitively man 
 agreed with the lower Primates in having a definite sexual season.* The fact is also 
 noted that conception (and, consequently, ovulation) can occur in the human female, 
 even although there may never have been a menstruation. The discussion of these 
 facts is inseparable from the question regarding the homology between menstruation 
 in the Primates and the procestrum in the lower Mammalia, a subject to which 
 reference has been made above. 
 
 TJie Atretic Follicle. — The sheep referred to in Experiment 6 furnished the only 
 unequivocal case of a follicle failing to rupture during an observed oestrus,! although 
 
 * Heape, in a later paper (1900), gives a very brief rdsumS of the evidence (derived largely from the 
 writings of Ploss and Westermarck bearing on this subject. Very recently Messrs. Annandale and 
 EOBINSON have kindly furnished me in advance with the following statement, which has not yet 
 appeared : — • 
 
 "One of the women of a Semang or aboriginal tribe in the Siamese Malay State of Jalor volunteered 
 the information that the women of her race only had children about March, that is, immediately after the 
 wet season, and that each woman normally had a child every year at this season, as long as she was 
 capable of child bearing." 
 
 I am indebted to Messrs. Annandale and EOBINSON for permission, to make use of this very interesting 
 information, which, so far as I am aware, is among the most direct evidence of a restricted sexual season 
 in a primitive race. (Cf. "Fasciculi Malayenses, Anthropology," Part I, now issued. — May 15th.) 
 
 t It is possible that the follicle referred to was not sufficiently advanced towards maturation to admit 
 of rupture, since it did not show any indication of atresia, so far as I observed. But this may quite well 
 be explained by the fact that the ewe was killed only on the day following the detection of oestrus, and 
 possibly but a few hours after the termination of that oestrus ; so that the follicle might not have begun 
 to degenerate. I am unable to state when the polar bodies are extruded in the case of the sheep. 
 
 Digitized by IVIicrosoft® 
 
72 ME. F. H. A. MARSHALL ON THE CESTEOUS CYCLE AND THE 
 
 the histological appearance of the atretic foUicle described in the preceding experi- 
 ment made it probable that it had undergone degeneration owing to insufficiency of 
 stimulus when in a condition for ovulation. Numerous cases, however, of follicles 
 becoming atretic before, and usually a considerable time before, attaining their 
 maximum development have come under my observation. 
 
 With the sheep, the follicle undergoing atresia can be identified by the following 
 characters, which distinguish it from the corpus luteum or discharged follicle : — 
 
 (1.) There is no indication of any rupture to the exterior. 
 
 (2.) The ovum, being retained in the follicle, degenerates, losing its regularly 
 circular shape as seen in section, becoming shrunken, and eventually disappearing 
 altogether. 
 
 (3.) The follicular epithelium, instead of hypertrophy ing, as in the formation of 
 the corpus luteum, degenerates, the chromatic substance at one stage appearing in 
 the form of fine points in the cytoplasm, much smaller than nuclei. The cells, in 
 various stages of degeneration, become scattered in the liquor, and often closely 
 resemble polynuclear leucocytes. I have never noted division among the epithelial 
 cells, nor deposition of fatty particles. Later, in atresia, the remains of the epithelial 
 cells become unrecognisable, finally disappearing altogether. 
 
 (4.) The connective tissue wall does not proliferate to form a network among the 
 epithelial cells. Generally there is no growth inwards from the theca until the 
 epithelial cells are in a very advanced state of degeneration or have entirely 
 disappeared. Eventually there is a loose ingrowth of connective tissue to fill the 
 cavity. 
 
 (5.) There may be a slight breaking down of blood-vessels early in atresia, so that 
 blood corpuscles are scattered in the follicular cavity. (I have never seen a blood- 
 clot in a sheep's atretic follicle, and only noted a few scattered corpuscles in a single 
 instance. ) 
 
 Where atresia occurs in very young follicles, they can obviously be distinguished 
 from developing corpora lutea by their small size. 
 
 I have made no attempt to obtain a complete series of stages illustrating the 
 degenerative changes which the atretic follicle passes through in the case of the 
 sheep. These changes have been described for various Mammalia (chiefly rabbits, 
 cavies, and other rodents) by Schulin (1881), Flemming (1885), Sohottlandee 
 (189], 1893), Henneguy (1894), Janosik (1896), and (for bats), van der Stricht 
 (1901), besides some other writers. My observations, so far as they go, are in general 
 agreement with their far more detailed accounts, which, in the main, are in accord 
 with one another. It would seem probable, however, that Henneguy (1894) has 
 mistaken for the membrana granulosa the thickened theca interna, since the ovum 
 described is obviously far advanced in atresia. ScHULiN (1881) and also Janosik 
 (1896) appear to regard the follicular epithelial cells as being actually converted into 
 leucocytes, which, as noted above, they frequently closely resemble when undergoing 
 
 Digitized by IVIicrosoft® 
 
FOinfATTON OF THE C'OllPUS IJTTEUAF TN THE SHEEI', 
 
 / '1 
 
 atresia. Flema[ing (iSSf)), du the other hand, denies the existence of leucocytes, 
 pointing out that noi^e occur in tlie theea, and S('HOTTL\xdrr (1891) clearly distiu- 
 o'uishes degenerating epithelial cells from leucocytes. SoHorrr.ANDER (1891) also 
 descinbes atresia as occurring by tatty degeneration as well as hy chromotolysis. 
 
 T^e presence of a degenerate ovum cannot apparently, by itself, be regarded as an 
 absolute indication of an atretic follicle, since Sobotta (1896, 1897) has recorded 
 instances of follicular discharge in the mouse and the rabbit, in which the ova were 
 accidentally retained in the cavity, the follicles nevertheless forming true corpora 
 lutea ; while van der Stricht (1901) describes a similar case of retention of the 
 ovum in Vesperufjo, where part of a ruptured follicle possessed the characters of a 
 young corpus luteum and another part those of an atretic follicle. 
 
 Atresia is commonly stated to be most frequent during pregnancy, but it occurs at 
 other times also (compare Janosik, 1896, and others"'), and according to my obser- 
 vations on the sheep, it is commonest among follicles that have attained to dimensions 
 from abo\it one-eighth to one-half of those of the matinee follicle. In a complete 
 series of sections through the ovary of a sheep killed in the month of August (during 
 the anrestrous period) I failed to discover a single atretic follicle, nor did I find any in 
 the ovary of a yearling lamb killed at about the same time. In another sheep's 
 ovary during anoestrum I have noted atresia. But my inability to find young 
 degenerating follicles oftener was probably in part due to my observations in ovarian 
 histology being mainly confined to material obtained during the sexual season or 
 verv early in pregnancy. But at these times to find young follicles undergoing 
 atresia is no very rare occurrence, for I have noted three or foiu' in a single ovary, or 
 even within the limits of a single section. 
 
 It is difficult to avoid associating the degeneration of developing Graafian follicles 
 in healthy ewes with detrimental environmental influences, and more especially with 
 scarcitv of food supplv, or of the kind of food supply necessary for breeding stock. 
 When occurring to any considerable extent, it must inevitably reduce the number of 
 ripe follicles in the sexual season, and thereby raise the barrenness percentage. 
 
 The question, then, concerning the causes producing follicular degeneration is not 
 without its practical side. To this subject I shall return presently. 
 
 Nnmher of Follicles Discharging at an CEstrous Period. — I give below details of 
 observations in regard to the number of follicles which discharge at a time {i.e., 
 during a single oestrus), the figures being derived from the examination of the ovaries 
 of fifty-five Scotch Black-faced ewes, or half-bred Black-faced ewes, in which ovulation 
 had occurred. The results were obtained either by counting the follicles shortly after 
 rupture (when the number which have recently discharged is usually perfectly 
 obvious), or by noting the number of corpora lutea in pregnant animals. All 
 doubtful cases (such as when it was Impossible by superficial examination to make 
 
 * SCHULIN (1881) says that atresia may set in at any stage, and even in tlic primordial follicle, 
 YOL. CXCVI. — B. L 
 
 Digitized by IVIicrosoft® 
 
74 MR. F. H. A. MARSHALL ON THE (LSTROUS CYCLE AND THE 
 
 out clearly how many corpora lutea were present) were passed over and not 
 included.* 
 
 Number of sheep m which a single follicle had ruptured at ovulation , 42 
 Number of sheep in which a single follicle in each ovary (i.e., two 
 
 follicles) had ruptured at ovulation ^ 
 
 Number of sheep in which two follicles had discharged in one ovary, 
 
 but none in the other, at ovulation ^ 
 
 Number of sheep in which two foUicles in one ovary and one in the 
 
 other (i.e., three follicles) had discharged at ovulation 1. 
 
 Total 55 
 
 Thus in less than 24 per cent, of the cases examined where ovulation had occurred, 
 was more than one follicle found to have discharged, t " 
 
 The number for Aviiich the percentage is calculated is probably too small to support 
 any general conclusions, but taken in conjunction with other considerations, is not 
 
 altogether unsuggestive. 
 
 Mr. He ape's records (1899) for 77,8,50 ewes belonging to 275 flocks of various 
 breeds, obtained for the year 1896—97, show an average return of 30-14 twin lambs 
 per 100 ewes. No similar records, so far as I am aware, have ever been made for 
 Scotch or other mountain sheep ; but there can be no doubt that with these the 
 percentage of twins is a good deal less than among Lowland or Down breeds. A 
 sheep farmer informs me from his own experience that from among fifty score of 
 Black-faced sheep not more than forty pairs of twins are to be expected, while among 
 Cheviots there may be 25 per cent, twins, and among half-bred Black-faced about the 
 same number, but tlie percentage may be higher. My informant states further, in 
 reference to both Cheviots and Black-faced sheep, that the twin percentage varies 
 considerably according to whether the ewes are kept on lowland pasture or on the 
 side of the hill, in the latter case there being practically no twins produced. State- 
 ments made to me by other sheep-breeders are in general agreement with these 
 rough estimates, but the twin percentage for pure Black-faced sheep is sometimes 
 said to be higher, the number of twins reaching 10 or 12 per cent. 
 
 Tt is of course obvious that where only one follicle ruptm-es at ovulation, only a 
 single lamb can be produced, and in \IeA\' of the percentage of follicles discharged (as 
 
 * The majority of these were from slaughter-house material ohtuined inrtiscriminatoly, l)tit I have also 
 counted in (lischargerl follicles (flescrilied elsewhere in this paper) from sheep that had licen under 
 observation, 
 
 t III the case of some of the slaughter-house material, 1 was unable to determine whether the sheep 
 were pure Scotch Black-faced, or crosses in which the character of the Black-faced breed was dominant. 
 But witli pure Black-faced sheep .alone, theie is evidence that the perceiLtagc in which two follicles 
 discharge is less, seeing that with the seven ewes (included) that had been under observation, only one 
 follicle had discharged in each instance, 
 
 Digitized by IVIicrosoft® 
 
.FORMATION OF THE CORPUS LUTEUM IN THE SHEEP. 75 
 
 noted) being scarcely, if at all, in exo^ss ol" the usual percentage of lambs produced* for 
 the breeds in question, the con\ erse of this statement is most probably also generally 
 true, x^nd since scarcity of twins is intimately associated with barrenness, as shown 
 by Mr. Heape (1899), and that habitually (i.e., constitutionally) barren ewesf are a 
 rarity, it is rendered exceedingly probable that infertility generally in sheep results 
 from the absence (or great scarcity) of ripe follicles ; that is to say, that in dealing 
 ■\\ith the causes of barrenness for the purposes of pi'actical sheep breeding, only a 
 single question need be considered, viz. , the scarcity of Graafian follicles available for 
 oA-ulation at " tupping time " {i.e., during the sexual sea.suii). 
 
 It is u common practice with scuue slieep-farmers to give the ewes un extra supjily 
 of food (siicli as turnips, cake, or corn, in addition to the pasture they ordinarily live 
 on) shortly before the " tupping season," so as to have them in good condition by the 
 time they are ready to take the ram, the object being to increase the percentage of 
 births in the succeeding lambing season. There is, however, some difference of 
 opinion as to the advantages gained by thus " flushing" the ewes, especially in those 
 cases where, after " tuioping time," they are placed on mere sustenance diet. Scott, 
 in an article on " The Practice of Sheep Farming," while advising breeders to have 
 the ewes in good thriving condition before tupping time, says that this " should 
 not be accomplished by a process of hurried and temporary stimulation." Although 
 deprecating "flushing," he admits that it may increase the number of births at the 
 ensuing lambing season (but " the extra number of lambs raised in any one year by 
 this system is on the average not very great"), while "from actual test" he is 
 convinced '" that ewes which have been flushed one year are never so prolific the 
 next." 
 
 When it is remembered that the ripening of the Graafian follicles is a process 
 which goes on not merely during "tupping time" and for a short period before but 
 throughout the '\'\'hole of the animal's lifetime, it becomes obvious that the practice of 
 "flushing" must lead in different individuals to diffei'ent results, depending upon 
 age, and the previous conditions to which the ewe has been subjected in regard 
 to food supply and environment, and so upon the extent to which follicles have 
 degenerated, the degrees of ripeness of the developing follicles, and consequently 
 the number which can be brought to maturity in the appi-oaching sexual season. 
 
 It follows also that "flushing" does not necessarily have the desired result even 
 in the following lambing season. That this is so is shown by two experiments 
 carried out for a very different object.]; under the direction of Girotj, by whom they 
 are described (1827). 
 
 * Allowance must be made for the possibility of one embrj'o aborting in a small proportion of cases. 
 
 t No case of a constitixtionally barren sheep has come under my notice. HeA-PE (1899) says that "it 
 is very generally conceded that 2 per cent, is a liberal allowance for constitutionally barren ewes, and that. 
 anything over 2 per cent, experienced is due to other causes." 
 
 I CtIROU's experiments were to determine the causes inllucnciiiy sex. Consequently in stating the 
 
 L 2 
 
 Digitized by IVIicrosoft® 
 
t6 MK. ¥. li. A. MARSHALL Ois THE CESTROUS CYCLE AND THE 
 
 In the first experiment, tAvo flocks during the tupping season were placed 
 respectively under different conditions, the one being supplied with more abundant 
 nutrition than the other. The following is the result of the subsequent lambing, 
 the ages of the dams Ijeing also given : — 
 
 Flocks under more favourable conditions. Flock under les.s favourable conditions. 
 
 Number of lamljs. 
 
 10 
 
 29 
 
 47 
 
 :■ 49 
 
 Age of dams. 
 
 
 Numljcr of lamlis. 
 
 Age of dams. 
 
 2 years 
 
 
 
 40 
 
 2 years 
 
 3 „ 
 
 
 
 45 
 
 ■J !! 
 
 4 „ 
 
 
 
 26 
 
 4 „ 
 
 5 „ aiid 
 
 over 
 
 
 26 
 
 5 „ and 
 
 
 Total 
 
 
 137 
 
 
 Total . . . 135 
 
 The number of sheep in each flock does not appear to be stated, but they are 
 described as having been of approximatel}' equal size. Each flock Avas provided Avitli 
 two rams. In the flock given more abundant pasture there Avere three cases of 
 tAvins, but none in the other. No cases of barrenness or abortion are mentioned 
 (this not concerning the subject of GiKOu's investigation); but since the flocks 
 are described as being composed about equally of eAves of all ages, it may perhaps 
 be inferred that some of tlie younger sheep among those given less abundant food 
 A\ ere barren. The total number of births in each flock is shown to be very nearly 
 equal. 
 
 In another experiment described by Gmou (1827) a flock of eighty-four eAves 
 Avas divided into equal parts, one-half being placed on much richer pasture than 
 the other during the sexual season. The result in regard to the number of lambs 
 produced Avas exactly the same in both instances, there being forty lambs born in 
 each half of the flock. Where the pasture Avas the richer there Avere two cases of 
 twins, Avith the other ludf one case of twins. 
 
 It is natural to assume that the fertility of young sheep is more sensibly affected 
 by an increased food supply and more favourable environment in the sexual season 
 than is likely to he the ease A^'ith the older ewes, since A\'ith the former there has 
 been less time and opportunity tlian A\ith the latter for the operation of preAuoiis 
 environmental influences, Avhich, if adverse, AAould haA-e letarded the ripening or 
 brought about the degeneration of such of the Graafian follicles as might otherwise 
 haA^e become mature. I have pointed out that the first of GiROu's experiments 
 affords some evidence of the truth of this deduction, while the opinion which 
 Mr. Heape (1899) says has not infrequently been expressed to him " that shearling 
 eAves are more liable to bairenness [presumably in unfavourable seasons] than older 
 
 number of lambs produced, GnxOU gives the number of male and female lambs respectively. As the 
 sexes of the lambn do not concern the subject of the present paper, I have added the numbers of males and 
 females lotcetlicr, and gi\ en only the totals. 
 
 Digitized by IVIicrosoft® 
 
i^ORMATIOX OF THE CORPUS LUTEUM IN THE SHEEP. 11 
 
 ewes," points in the same direction, altliougli the returns submitted to him do not 
 show any proof of this except for the Dorset l)reed. For these " there is some 
 evidence that in unfa>'(iurable tupping seasons, during great heat and scarcity of 
 green food, the young ewes are more hi-celv to he affected than tlie older ewes, 
 and less likely to he in a satisfactory hreetling condition." 
 
 Of miore doubtful bearing on this subject is Scott's statement, ah'eady cj^uoted, that 
 ewes from which an extra number of laml)s ha\e been raised by Hushing one year are 
 less prolific at the subsequent lambing, or tlie fact referred to in the section of this 
 paper on the " QCstrous Cycle," that those Limestone slieep which produce lambs twice 
 during one year a. re barren in the \'ear following. Such facts as tliese are usually 
 ascribed to general strain on tlie reproductive system, but when this sti'ain is said to 
 cause sul>sec|uent barrenness (as distiiict from abortion or "slipping lamb''), I would 
 suggest that the barrenness is due to a premature but partial ripening ol' the 
 (.TraaHan follicles succeeded by an unusual amount of follicular atresia. ''•■' 
 
 The general practical conclusion to be drawn fiom the above considerations is that 
 it is better to keep the ewes in fair thriving condition, as continuously as possible, 
 than to submit them to an unnatural process of stimulation by. " flushing " shortly 
 before and during the sexual season, while maintaining them on mere sustenance diet 
 at other times of the year. 
 
 There is another point of some practical importance which deserves notice. It is 
 sometimes asserted by breeders that the fertility of sheep is affected by the ram 
 employed {e.g. , that the production of twins or triplets in place of single lambs is in 
 some way due to the use of particular rams). In the light of the facts stated above 
 regarding the production of ova, and remembering the far greater reproductive capacit}' 
 of the ram as conapared \\\\\\ the ewe, these assertions (excepting where they can be 
 shown to apply to the causes of abortion) may be disregarded ; and as a practical matter. 
 the question of fertility (at least ordhiarily) may be considered as one aftecting the 
 ewe alone. At the same time, it is obvious that to ensure copulation occurring with 
 every e^^'e, the proportion of ewes to rams nmst not be too high, while there is 
 evidence that the sexual activitv of the ram is liable to vary, both individually and 
 according to the breed. 
 
 The Border Leicester and Scotch Black-faced rajns afford an example of this 
 variation. It is a common practice in Scotland ^\\\\i breeders of crosses between these 
 two varieties of sheep to keep Black-faced rams to ser\'e those Black-faced e\\es which 
 in the same sexual season have failed to become pregnant to a Border Leicester ram, 
 
 * I have never met witli a case of cestrua occurring with a ewe in the absence of protruding or very 
 recently discharged follicles. Whether or not oestrus does ever occur with sheep without the possibility 
 of ovulation («.e., with barren sheep) I am unable to say. But to assume that it does not ordinarily do 
 so is not to assert that cestrus depends upon ovulation, since the two phenomena may both be results of 
 the same or similar causes. In some Mammalia, as has liecn pointed oijt l)y Heape, o\ ulation and cestrus 
 can occur quite iiideiicndcntlj'. 
 
 Digitized by IVIicrosoft® 
 
78 MK. ¥. H. A. MARSHALL ON THE CESTROUS CYCLE AND THE 
 
 the failure being merely the I'esult of the latter not copulating, and not due to 
 infertility on the part of the e^^'e,s. I am informed that ^\■ith such cases, in the 
 lambing season, as many as half the lambs produced may be pure Black-faced, and 
 from e\N-es which would have tailed to bear, had the work been left entirely to 
 
 Leicester rams.* 
 
 The statement that Suffolk and Wensleydale ewes are more prolific with rams of 
 their breeds than with other rams (Heape, 1899) may perhaps be similarly 
 
 explained. 
 
 On the other hand, with Dorset Horns, Mr. Heape describes the ewes as being 
 less barren with Hampshire Do\vn rams than with Dorset Horn rams. Referrmg 
 to a general instability of the reproductive system of Dorset Horn eAves, and 
 their need for some special stimulus, he appears to suggest that this is supplied 
 by the use of a Down ram. I think that it is more probable that what m these 
 cases is described as barrenness is in reality ^ery early abortion, an unavoidable 
 element of error, vs'hicli Mr. Hea.pe recognises as present in his statistics for sheep 
 generally, (though he states that he belieAes the error is, as a rule, small) being in 
 the case of the Dorset sheep somewhat larger than elsewhere in the returns given. 
 That this is the true explanation is rendered the more probable by the facts also 
 recorded bv Mr. Heape, " that, of all breeds, the Dorset Horns appear to be most 
 liable to abortion," and further, that in some cases, " Dorset Horn ewes Avhich slip 
 lambs got Ijy Dorset Horn rams will bear lambs got by DoAvn rams." I would suggest 
 that in this case the cause of abortion lies, not so much with the ewe, as in want of 
 vitality in the embryo, Avhich A\-hen aborted early is absorbed in the uterus, escaping 
 the notice of the shepherd ; and that A^diere the Dorset Horn sheep have been served 
 hv Down rams, there is less abortion as a result of increased vigour on the part of 
 the cross-bred young. 
 
 The Fonnation oftJie Corpus Lateuin. 
 
 The ages of the corpora lutea in this investigation were reckoned from coition, 
 
 or ■^^'hel•e coition did not oi' \\'as not known to Ivayh taken place from the time 
 
 when a'strus was observed, the slieep being killed at various intervals of time 
 
 afterwards. The facts recorded in the previous section of this papei' indicate that 
 
 ovulation only takes place during the oesti-ous periods (or veiy possibly in some 
 
 instances in the prooestrous periods which immediately precede them), while the 
 
 observed relation, as described belo\\', betAveen the condition of development of the 
 
 corpus luteum and the interval that had elapsed IjetAxeen the oestrus noted and the 
 
 killing of the sheep in question, is' by' itself strong' c\idence that the two 
 
 phenomena occur together. But the method adopted of reckoning tire' ages of the 
 
 deA'eloping corpora lutea is inevitably a somewhat arbitrary one, since, as already 
 
 * The failurt) of the Leicester rams here must l3c piirtly due to their being less suited- to the Scotch 
 ulimiitu ill the hitter part of the sexual season of Black-faced ewes. 
 
 Digitized by Microsoft® 
 
FOT^MATinX OK TITK COltPUS TJ^TEUM T\ TITE STTEEr. 7',) 
 
 vemarkeil, thero is no (lefinitc time (Uiriiii;' n\sli-us iit which DNulntinii occiiis, while an 
 (ostroiis period may extend over seNcnd lionrs. ( !onsequentlv two recently discharged 
 follicles, although of the same age, according to the method of reckoning, niaA' 
 present slight!}' different degrees of (U'velopment * 
 
 I proceed to descrihe the stages which I haxe ohtalned in the formation of the 
 sheep's C(^rpus luteum. When coition A\-as known to hax^e taken place, the fact is 
 recorded. 
 
 The corpus luteura ohtained from an animal killed fieven liourx after coition differs 
 from the unburst follicle in its size and in the fact that the ovum and discus 
 proligerus have heen discharged. It is a little more than half as large as the mature 
 follicle, and consecpiently does not protrude from the surface of the ovary. On the 
 other hand, the point of rupture is situated somewhat in a depression. Very little 
 blood is found remaining in the cavity of the follicle, but a few scattered corjjuscles 
 are seen in the membrana granulosa, which was not discharged with the ovum and 
 discus proligerus. The blood corpuscles are apparently derived from vessels, the 
 walls of which have broken down, not merely near the point of rupture of the follicle, 
 but to a less extent around the whole theca interna. The greater quantity of the 
 blood corpuscles set free from the vessels at the rupture of the follicle had been 
 discharged to the exterior, forming a stain, in which the blood had not yet clotted, 
 on the surface of the ovary in the region of the depression communicating with the 
 follicle's cavity. This latter contains a fluid resembling in all respects tlie liquor 
 folliculi. The membrana granulosa at this stage of development is approximately 
 twice the average thickness of that of the undischarged ripe follicle, some of the cells 
 liaving increased in size, while others, especially those nearest to the periphery, 
 retain the characters of the original follicular epithelial cells. Evidence of recent 
 mitosis among these cells is not uncommon, though not so frequent as Avith tlie 
 epithelial cells of the Graafian follicle. In this specimen there is no sign of any 
 growth inward of connective tissue from the theca interna, which, except at certain 
 points where blood-vessels had ruptured, did not so far appear to present anv 
 change. 
 
 In another seven-hour stage, the development of the corpus luteum is in a slightly 
 more advanced state. Ingrowths, at this period resembling little bud-like projec- 
 tions arising from the theca interna, are seen on the inner side of that layer, and 
 although blood-A'essels are also beginning to grow inwards with the connective tissue, 
 the proliferation of the theca cell is not as yet very great. The anastomosis of 
 fusiform cells, which arises from division of the cells of the theca, and which eventu- 
 ally comes to surround neaidy every one of the follicular epithelial cells, at this stage 
 has not begun. The membrana granulosa already shows considerable hypertrophy, 
 but still retains the irregularity of arrangement and want of cohesion which results 
 from certain of its component cells having shifted their position at the time of 
 
 * See p. G8, 
 
 Digitized by Microsoft® 
 
80 ME. F. H. A. >rA?v«HALL ON THE CESTROUS CYCLE AND THE 
 
 rupture. This irregularity is more marked than ^^itli the seven-hour stage previously 
 described, and may to some extent result also from the ingrowing connective tissue 
 displacing certain of the epithelial cells from the positions which they had assumed 
 after the discharge of the follicle. As a consequence of the ingrowth, the membrana 
 propria no longer exists as an intact line. Wandering cells, free in the connective 
 tissue, are found to occur not infrequently, hut these do not appear to have been 
 extra vasated during the rupture of the vessels. In the case under consideration, 
 practically no blood corpuscles are found remaining in the follicle's cavity or 
 distributed among the epithelial cells, but in this instance, as in all other examples of 
 recently ruptured Graafian follicles in the sheep that have come under my observa- 
 tion, a considerable quantity of blood was discharged to the exterior. A section 
 through the wall of this follicle is figured on Plate 8 (fig. 8), where it may be 
 compared "s^-ith a section through the wall of an undischarged follicle (fig. 7). 
 
 Tt may here he noted that the vessels of ovaries containing recently discharged 
 follicles may present signs of congestion apart from those in the region of the young 
 corpus luteum. There is also evidence of increased blood supply to the ovaries 
 during the j)rooestrum, i.e., before the occurrence of ovulation, but I have never seen 
 any breaking do-\An of blood-vessels Avithin the ovary, except in connection with a 
 newly ruptured follicle. 
 
 In a discharged follicle obtained sixteen hours after coition (fig. 9, Plate 9) the 
 connective tissue ingrowtli is carried further, and cells, generally fusiform in shape 
 and arising by mitotic division from those of the theca ingrowths, are beginning to 
 penetrate between the epithelial cells. The increase of connective tissue is accom- 
 panied by a corresponding increase of vessels which take their origin from those of 
 the theca wall, while the walls of vessels which had given way appear to have been 
 recuperated. In this case also, extvavasated blood corpuscles within the cavity of 
 the follicle or free in the tissue are rare. The communication with the exterior 
 where the follicle had ruptured is stdl widely open, and here red corpuscles and 
 follicular epithelial cells are found in considerable quantities. The latter are either 
 isolated or arranged in small groups, and alread}- some of them present signs of 
 degeneration, ^vhile many are slightly hypertrophied. Isolated epithelial cells 
 presenting similar characters are numeroiis everywhere in the fluid- containing cavity, 
 which is far smaller than that of the mature follicle. These cells had no doubt 
 mostly formed part of the discus proligerus, but certain of them might have belonged 
 to the membrana granulosa. The vast majority of the granulosa cells, however, still 
 maintain their position, but have further Increased in size. The average thick- 
 ness of the epitlielial ^vall of the corpus hiteum at this stage is not shown in 
 the figure, which \\as drawn at a point where this layer is much more reduced than 
 elsewhere in the same follicle, the object being to represent a high-power mag- 
 nification of a section tln-ough the complete follicular wall from the cavity to the 
 
 Digitized by IVIicrosoft® 
 
FORMATION OF THE CORPUS LUTEUM IN THE SHEEP. 81 
 
 The most remarkable characteristic of the corpus luteum of this stage of develojD- 
 meut is the number of leucocyte's, which I have found to be far more abundant than 
 at any other period. These leucocytes occur both in the connective tissue and also 
 free in the follicular cavity. Many of them, at any rate, undergo degeneration, 
 when they resemble degenerating follicular epithelial cells, which, as already noticed, 
 frequently occur isolated in the caAit}'. In the case of the atretic follicle, degene- 
 rating epithelial cells have been mlstaktm for l(;ucocytes, as remarked above (p. 72), 
 while in certain later stages of the formation of the corpus luteum, 1 have found it 
 difficult, if not impossible, to distinguish then) -when In a more advanced state of 
 degeneration. 
 
 The corpus luteum of iwenty-four hours has undergone further changes. Its 
 Increase in size is well marked, its dimensions now approaching those of the mature 
 follicle. Its shape is irregular, while Its walls have become more folded. The 
 central cavity Is smaller, but still communicates with the exterloi-, where it opens out 
 Into a cup-shaped depression (in the surface of the ovary, from which the corpus 
 luteiun now appreciably protrudes. The epithelial wall of the cavity is at this 
 period at least twice the thickness of that of the seven-hour stage, this Increase being 
 due for the most part to the simple hypertrophy of the individual cells composing it, 
 these appearing In section to be at least two oi' three times the size of the membrana 
 granulosa cells of the undischarged Graafian follicle. Division may, however, still 
 be occasionally observed among the epithelial cells. But the thickness of this layer 
 of the follicle's wall is also increased by the ingi'owth of connecti\'e tissue, which b}' 
 this time has become considerable. Groups of epithelial cells have become surrounded 
 by a network of fusiform cells, derived from strands of ingrowing connective tissue, 
 which give the young corpus luteum a radial appearance. The inter-epithelial 
 connective tissue, consisting mainly of fusiform cells, and the hypertrophled epithelial 
 cells are shown in fig. 10 (Plate 8), which represents a part of a section through the 
 wall of a ruptured follicle of this stage. But although the connective tissue element 
 of the corpus luteum of the sheep is provided chiefly by the proliferation by mitotic 
 division of the cells of the theca interna, it Is in part derived from the more fibrous 
 theca externa, from which layer strands of cells, often in close connection with those 
 of the inner layer, are at this period beginning to grow inwards among the epithelial 
 cells. But the inner theca layer as such has to a great extent disappeared, having 
 been used up in the formation of inter-epithelial connective tissue. It is to be noted, 
 however, that the line of demarcation between the two thecse comprising the 
 connectiA'e tissue wall of the follicle is not always \\ell defined, either at the stage 
 now being described or at other stages in the formation of the corpus luteum, while 
 the boundary betAveen the theca externa and the ovarian stroma is frequently still 
 less definite, particularly In certain of tlie later stages. Extravasated blood 
 corpuscles scattei-ed in small patches and streaks occur in the twenty-four-hour stage, 
 as in the earlier stages, but are not very numer(nis, there being no blood-clot within 
 
 VOL. CXCVL— B. .M 
 
 Digitized by IVIicrosoft® 
 
82 MR. F. H. A. MARSHALL ON THE CESTROUS CYCLE AND THE 
 
 the cavity. Leucocytes, though present, do not appear m such numbers as in the 
 sixteen hour stage previously described. 
 
 In the corpus luteum thirtij hours after coition the inner theca layer has all but 
 vanished, while the inorowth of strands from the outer theca is further advanced. 
 The follicular epithelial cells are, on the average, slightly more hypertrophied, and 
 the connective tissue netA\-ork ^vhich in the previous stage had already begun to 
 inclose some of them, has become more elaborated. The growth of connective tissue 
 is accomjjanied by the increase in the number of blood-vessels. Red corpuscles 
 extravasated at tlie time of discharge are more numerous AA'ithin this follicle than in 
 the cases of certain of the other stages, and are especially numei'ous in the proximity 
 of the point of ruptui'e, ^vhich is still ojDen, and communicates witli the fluid- 
 containing cavity. Fig. 1 1 illustrates a section through the opening of the follicle, 
 showing the depression on the surface of the ovaiy, and the beginning of the passage 
 leading to tlie interior of the follicle. Islands of follicular epithelial cells are seen 
 to be present near the aperture, and abundant extravasated blood corpuscles, as well 
 as leucocytes ; also newly formed blood-vessels, which have made their appearance 
 Avith the ingrowth of the connective tissue, as shown in the figure. In some places 
 I have found it a matter of great difficulty to distinguish between corpuscles extra- 
 vasated at the time of the follicle's rupture, and corpuscles in thin-walled vessels 
 which take theii' oiigin fi-om those ah'eady existing in the theca wall. The bulk of 
 the blood present, however, is clearly extravasated. Fig. 13 is a niicrophotograph 
 of a transverse section of the ovary containing the ruptured follicle here described, 
 and gives some idea of its size in relation to the rest of the ovary, besides showing 
 the aperture of tlie follicle, aud the communication y\'\t\\ its already considerably 
 obliterated cavity. (Plate 10.) 
 
 The epithelial cells of the corpus luteum of about fifty Ilouvs as seen in transverse 
 section are four or five times the size of those of the undischarged follicle. Mitotic 
 division is very rare among them, but evidence of it may still occasiojially be observed. 
 Proliferation of the connective tissue cells continues to take place, chiefiy in the 
 direction of the central cavity, which has become smaller. Leucocytes, some of 
 which are apparently degenerate, are to be seen among the epithelial and fusiform 
 cells, as M-ell as free red corpuscles. A few degenerate epithelial cells also occur 
 isolated in the cavity. The corpus luteum as a AAdiole presents a radial ap])earance, 
 which is emphasised by the ingrowth of large strands of connecti^-e tissue taking 
 their origin from the theca externa. The iniier theca layer, as such, has entirely 
 disappeared. 
 
 The corpus luteum of sixi?/ /lowrs is larger than the ripe follicle. The epithelial 
 cells are still increasing in size by simple hypertrophy, but I have not observed any 
 case of division, the connective-tissue cells are dividing in all directions, so that 
 nearly every epithelial cell has become surrounded by an anastomosis of fusiform 
 cells. Blood-vessels, now of \-cry considerable size, moy l)c seen in many i)laces in 
 
 Digitized by Microsoft® 
 
■rOEMATION OF THE CORPUS LUTEUM IN THE SHEEP. 83 
 
 the epithelium, in ci^uneetion with the conuecti\»^-tis,sue ingrowth, and especialls' 
 near the periphery, where they arc otiolosed hy large atrfuuls of oells arising from the 
 theca externa. The point wliere the follicle liad ruptured is still recognisable at tho 
 surface of tlie ovaiy, Init the passage eonmiunicating A\ith tho interior is almost 
 completely obliterated. The cavity itself is A-ery much reduced, and, though still 
 holding a fluid, no longer contains degenerating leucocytes or isolated epithelial cells, 
 all traces of which seem to have entirely disappeared. Extrav asated blood corpus- 
 cles, also, are absent in the specimen in question. 
 
 The process of obliteration of the cavity is eftected, partly by the continued hyper- 
 trophy of the membrana granulosa cells and the increase of the inter-epithelial 
 connective tissue, and partly by the formation of a wall of connective tissue within 
 the follicular epithelium. This wall is formed of elongated cells arising from 
 the continued ingrowth of the inter-epithelial connective tissue, and at the sixty- 
 hour stage consists of a perfectly regular layer, about six cells deep, completely 
 enclosing the cavity of the follicle. Fig. 12 illustrates part of a section through 
 the wall of the follicle of this stage, and shows the inter-epithelial connective 
 tissue, and the inner connective-tissue layer lining the edge of the now much 
 obliterated cavity. (Plate 10.) 
 
 In the corpus luteum three days after oestrus, the process of formation has been 
 carried still further. The point of rupture is only noticeable on the surface, where 
 there is still a shallow depression, but there is no communication with the interior, the 
 cavity having been all but filled-in by the further growth of connective tissue, the 
 fluid which it contained being absorbed. The follicular epithelial cells as they 
 appear in section are on the average about five or six times as large as those of the 
 unaltered membrana granulosa from which they were derived, but their actual 
 volume must obviously have increased by a much higher multiple. Fatty particles 
 have already made their appearance in a very few of them which have begun to 
 assume the character of true lutein cells. Meanwhile the inter-epithelial connective 
 tissue has become still more finely distributed. 
 
 The discharged follicle ybt/j" days after cestrus has already begun to assume the 
 chief characteristics of the fully formed corpus luteum. The place where rupture 
 had occurred has almost ceased to be visible, while the remains of the cavity are only 
 to be found in a few sections passing through the centre. Fatty particles are some- 
 what more numerous in certain of the epithelial cells, which are approximately six 
 times as large as the membrana granulosa cells, as they appear in section. The whole 
 structure is a very little larger than the corpus luteum in the three-day stage of 
 development. Extra vasated blood corpuscles and wandering cells no longer occur, 
 but the supply of blood-vessels is rich. The proliferation of connective tissue is still 
 going on, especially in the form of ingrowing strands from the theca externa. 
 
 Certain further stages in the formation of the corpus luteum, the ages of which 
 were not known, show a more advanced degree of development. The dimensions 
 
 M 3 
 
 Digitized by Microsoft® 
 
84 ME. F. H. A. MARSHALL ON THE CESTROUS CYCLE AND THE 
 
 of these were greater than with the stages described above, and were considerably 
 in excess of those of the ripe Graafian foUicle. The size of the corpus luteum is, 
 however, no sure guide to its age, for I have observed t^vo in the same ovary and ot 
 the same degree of structural development but of slightly different dimensions. In 
 the more advanced stages, the epithehal cells continue to hypertrophy until, as seen 
 in section, they have attained more than six times their original size. Meanwhile 
 the number of fatty particles in them has increased, and they have become converted 
 into those lutein cells which characterise the fully developed corpus luteum. 
 
 Other accounts of the Formation of the Corpus Luteum. — The first to discuss the 
 mode of development of the corpus luteum was vON Baer (1827), according to whom 
 the structure resulted from changes in the connective-tissue wall of the follicle, the 
 membrana granulosa not ^participating in the process. Subsequently Patekson 
 (1840), who described the corpus luteum of the sheep among other annuals, 
 attributed its formation to changes in the blood coagulum left within the cavity of 
 the follicle after its discharge. About the same time, Bischoff (1842) adopted the 
 \ iew that the lutein cells were derived from the membrana granulosa. Pateeson's 
 I'lypothesis never gained any general support, but the theory of von Baer, which 
 supposes the corpus luteum to be an entirely connective-tissue structure, and that of 
 BiscHOFF, deriving the lutein cells from the follicular epitheliimi, have, with com- 
 paratively slight modifications, gained numerous adherents. 
 
 To SoBOTTA belongs the credit of being the first to adequately aiDjily the experi- 
 mental method to the solution of the problem, and with the publication of his 
 investigation, the controversy regarding the mode of formation of the corpus luteum 
 entered upon a new jjhase. I do not propose to cite afresh the numerous authors 
 who have written upon this question prior to the publication of Sobotta's first 
 paper (1895) on the corpus luteum of the mouse, since an admirable resume of the 
 earlier as well as the later literature of the subject has been given by Sobotta him- 
 self in Merkel and Bonnet's ' Ergebnisse der Anatomie und Entwickelungsge- 
 schichte,' (1899), to which the reader is referred. 
 
 Sobotta, in obtaining the developmental stages of the corpus luteum of the mouse 
 (1895, 1896), and the rabbit (1897), collected his material exclusively from animals 
 kept under observation, and killed at stated periods after coition, so that Sobotta 
 adopted the same experimental method as that employed by the present writer in 
 investigating tlie corpus luteum of the sheep. His researches resulted in confirming 
 Bisciioff's hypothesis that the lutein cells are derived from the foUicular epithelium, 
 while the inner theca wall of tlie follicle is showai by Sobotta to give rise only to an 
 inter-epithelial network, and to a part of the connective tissue, which fills in the 
 cavity of the ruptured follicle. 
 
 Since the result of my investigation f.n the corpus luteum of the sheep agrees 
 essentially with that obtained by Sobotta for the corpora lutea of the mouse (1896) 
 and the rabbit (1897), it will be convenient to refer merely to the more important 
 
 Digitized by Microsoft® 
 
FOriNfATTON OF THE CORPUS LTTTElT^r TN THE SHEEP, 85 
 
 differences liotAvoon his nrcoiuil and mine. Some of tli(.'K(( diffcironccs liase been 
 mentioned by Sobotta in Ins nidst icccnt paper (190'J) while fonimcntino- on niv 
 ' Preliminary ( 'omnmn lent ion ' ( 1 901). 
 
 Accordino- to Sohotta's descriptions, tlie theca externa takes no pint in tlie 
 formation of tlie corpus hxtemn either in the mouse or in tin; rabljit, the comiecti\e- 
 tissue proliferation bein<;- dt'iived (Mitirely from the theca interna, in this i-espect differing' 
 from the corpus luteum of the sheep. But I am able to confirm Sobotta's statement 
 in the case of the mouse, for in sections through corpora lutea belonging to certain of 
 the more advanced stages, kindly lent me by Dr. J. PI. Ashwobth, I ha^•e been 
 unable to recognise any ingro^\th from the theca externa. The nature of the 
 connective-tissue proliferation in the early stages also differs somewhat in the two 
 accounts, as a comparison between my figures and those of Sobotta will show, 
 the ingrowths from the theca interna arising generally as bud-like processes in 
 the case of the sheep's corpus luteum. Another point of difference relates to the mode 
 of obliteration of the central cavity, which according to Sobotta is in part fdled in b\' 
 the transformation of leucocytes into star-shaped connective-tissue cells, and not 
 merely by the further ingro'wth of connective tissue arising from the proliferation of the 
 cells of the theca wall. In the development of the sheep's corpus luteum leucocytes 
 are present in great numbers in the early stages, and especially in the sixteen- 
 hour stage, as described above, but many of them are in a degenerate condition, 
 while in subsequent stages, prior to the filling of the cavity, leucocytes are almost 
 entirely absent. In the sixty-hour stage, the process of fiUing-in of the cavity 
 has started, and is seen to occur by the formation of a regular wall of cells, arising 
 from the inter-epithelial connective tissue. (Fig. 12, Plate 10.) 
 
 Sobotta describes the follicular epithelial cells in the mouse and rabbit, as 
 becoming transformed into lutein cells by a process of simple hypertrophy, mitotic 
 division almost entirely ceasing after the follicle's rupture. My own observations on 
 the sheep's corpus luteum to a very great extent confirm Sobotta's account, but I 
 have shown that mitotic division among the epithelium may continue with decreased 
 frequency in the earlier stages.''' 
 
 Other differences between Sobotta's account and mine relate to the absence, in 
 every case, of a blood coagukmi within the sheep's ru2Dtured follicle, and to the fluid 
 found in the cavity of the latter, this varying somewhat in quantity, and resembling 
 the licj^uor of the undischarged follicle. Sobotta shows that the place \A'here the follicle 
 ruptures heals up very rapidly in the case of the mouse, while in the rabbit, as in the 
 sheep, it remains open for a considerable period. The tendency to greater irregularity 
 in shape with the developing the corpus luteum of the sheep, as compared 
 especially with the mouse, though noticeable, is a point of no significance. 
 
 The development of the corpus luteum in the rabbit has been investigated also by 
 
 * Sobotta (1897) describes a single instance of mitotic division in n, folliculav cpithcli.'d cell of a 3'oung 
 corpus lutexim of a rabbit. 
 
 Digitized by IVIicrosoft® 
 
SG MR. F. H. A. MARSHALL ON THE O^sTROUS CYCLE AND THE 
 
 HoNORK (1900), who obtained a series of stages similar to those procured liySoBOTTA, 
 tlie animals heing Icilled at intervals after copulation, as in Sobotta's researches. 
 HoxoRl^'s conclusions are almost in entire accord with those of Sobotta, but the 
 more i-ecent investigate r describes the theca externa as taking part in the growth 
 of the inter-epithelial connective tissue, in this point differing from Sobotta, but 
 agj-eeing Avith mv observations on the formation of the corpus luteum in the sheep. 
 On the other liand, according to Hoxore, the -whole of the theca interna is not used 
 up in giving rise to the connective-tissue anastomosis, whereas according to the 
 investigations of Sobotta and myself this Lxyer entirely disappears in comparatively 
 early stages of development. 
 
 The observations of Stratz (1898) upon developing corpora lutea in Tarsius, 
 Tupaiit, and Sorex are in general agreement with Sobotta's desci'iptions. Van deu 
 Strtcht's account (1901) of the formation of the corpus luteum in bats (Vesperugo, 
 Plecotus, and Vespertilio), however, presents differences of some importance. This 
 author agrees with Sobotta in describing the follicular epithelial cells as persisting 
 and giving rise to the lutein cells, but he adopts the view that all the lutein cells are 
 not formed in this way, a few being derived from interstitial cells of the theca 
 interna. Vax der Stricht also differs from Sobotta in describing the not infrequent 
 occurrence of mitotic division, in addition to simple hypertrophy, among the lutein 
 cells, and notes such division in the develo^sing human corpus luteum, as well as in 
 tliat of the bat. In other respects tan der Stricht's description is confirmatory of 
 that of Sobotta. A special point of interest in van der Stricht's account is the 
 very early appearance of fatty particles in the follicular epithelial cells of the bat's 
 corpus luteum. 
 
 KoPSCH, at the meeting of the " Anatomische Gesellschaft " at Bonn, exhibited 
 some preparations of corpora lutea 1)elonging to three stages of development (three, 
 six, and ten-day stages) from the so^^-, confirming the results of Sobotta, Honore, 
 Stratz, and the present writer regarding the origin of the lutein cells (Sobotta, 
 1902). Sobotta (1899) states also that Bonnet has to some extent investigated 
 the formation of the corpus lute\mi in tlie dog, and obtained similar results. 
 
 Kreis (1899), writing on the development of the " cor^ms luteum spurium "'" or 
 '^ corpus luteum menstruationis," \\hich he appears to regard as distinct both from the 
 corpus luteum verum {i.e., the corjxis luteum of pregnancy) and from the atretic 
 follicle, describes the epithelium as being retained in the follicle, but adds nothing to 
 the discussion on the origin of the lutein cells. 
 
 Belloy's account (1899) of tlie formation of tlie corpus luteum in the rat and 
 the cavy differs from all others in descril)ing an acti\'e multiplication of the 
 fblhcnlar epithelial cells, which lie seems to me to have confused witli the iuffrowinc. 
 
 * The roi'pus luteum ,,f pregnancy, or corpus luteum verum, and the corpus luteum spurium or 
 menstruationis arc, by derivation, necessarily identical structures, and are generally so regarded. ' 
 
 Digitized by Microsoft® 
 
FORMATION OF THE CORPUS LUTEUM IN THE SHEEI'. 87 
 
 connective-tissue cells. At aiiv rate, Bl';L^(>^■ does not a])pt'ar to recoi^nise tlie nature 
 of the connective-t issue proliferation. 
 
 Mr. Hkapk (1897), Avithout discussing the origin of the hitein cells, states that the 
 increase hi the size of the wall of the discharged follicle in monkeys is chie to tlie 
 hypertrophy of its cells, and not to an increase^ in their uuinlter, this I'esult having 
 heen arrived at quite independent! v of Sot.oti'a's researcli. 
 
 IvABL (lbi;»8V who has also written on the iorination of the (•or])Us hitenju, is of 
 opinion that the hitein cells arise partly from the fdlicular ejtithelium and partly 
 from the connectiye tissue wall. 
 
 The autlmrs cited aho\e have all t() a. greater or less extent confirmed Sobotta's 
 account of the formation of the corpus luteum. A number of investigators, however, 
 since the puhlication of Sobotta's work have adopted the directly opposite hypothesis, 
 that the corpus luteum is an exclusively connective-tissue structure. Among them 
 KoELLiKEE. (1898) and His (Koelliker, 1898), at the meeting of the " Anatomische 
 Gesellschaft " at Kiel, made statements upholding their previously formed conclusions, 
 while Nagel (1896, see also Nagel, 1899) has also described the lutein cells as 
 ari.sing from the conuecti^•e tissue of the theca wall. The same opinion is likewise 
 maintained in a recent discussion of the question by Paladino (1900). 
 
 Clark (1898), after investigating the formation of the corpus luteum in the sow 
 and in the human subject, has convinced himself of its entirely connective-tissue 
 origin, expressing the hope that he had obtained I'esults which could leave no 
 doubt regarding its juode of development ; while Doering (1899), also \\riting on 
 the corpus luteum of the sow, claims to have confirmed Clark's account, Others 
 who have recently taken the same view are Buhler (1900), Wendeler, and Stockel, 
 the last author stating that he is forced to the conclusion that the lutein cells are 
 deiived from the theca intei'na. 
 
 The descriptions of all these authors have been dealt with by Sobotta (1898, 1899, 
 1902), so that it is unnecessary for me to do more than to refer to their papers, and 
 to state that I am substantially in agreement with Sobotta's criticisms. The results 
 arrived at by Nagel, Clark, and those who uphold the connective-tissue origin of 
 the corpus luteum, are due largely to confusion betAAeen discharged and atretic 
 follicles, and betAA'een the young and somewhat later stages in the development of the 
 discharged follicle, this confirsion resulting in great measure from the absence of 
 method in procuring material. 
 
 As already shoAvn, my own accoujit of the sheep's corpus luteum agrees essentially 
 A\ith that given by Sobotta for the mouse and the rabbit, and by Honore for the 
 rabbit, the material in each investigation having been obtained according to the 
 experimental method. In so largely confirming their accounts, my results, as 
 Sobotta (1902) points out, in commenting on my preliminary paper (1901), indicate 
 the falsity of the suggestion of His that the mode of development of the corpus luteum 
 in the larirer mammals radically differs from Avhat occms in the mouse and the rabbit. 
 
 Digitized by IVIicrosoft® 
 
88 ME. F. H. A. MARSHALL ON THE (ESTEOUS CYCLE AND THE 
 
 SoBOTTA alludes also to the rapidity of the process of formation of the sheep's corpus 
 luteum, a rapidity which is all the more remarkable owing to its resulting, to so great 
 an extent, from the simple hypertrophy of certain of its constituent cells. A parallel 
 to this rapid cellular hypertrophy may perhaps be found in the development of the 
 uterine outgrowtli in the Nematode Splmrularia, the cells of which are described 
 (8H1PLKY, 1896) as increasing in size but not in number, as is so largely the case with 
 those of tlie follicular epithelium, which plays such an important part in the formation 
 of the corjjus luteum. 
 
 Siuiuaary and General Conclusions. 
 
 In Scotch Black-faced sheep the duration of the dioestrous cycle varies from about 
 two to three weeks, according to the locality, and the same is probably the case Avith 
 sheep belonging to other breeds. The length of the sexual season and the number of 
 recurrent diwstrous cycles in a single season are also subject to variation, there being 
 ;i perfect gradation fi'om the dioestrous condition of the Highland Black-faced sheep 
 to the extreme of polyoestrum reached by certain Merinos. The difference is in part 
 related to the food supplv and environment, but increased sexual power is also a 
 racial characteiistic, as is shown, for example, by the Dorset Horn and Limestone 
 sli(.>ep. and by the sheep of Patanl. Wild sheep are probably for the most part 
 moncestrous, ^^•ith a tendency to polyoestrum under favourable conditions. The same 
 may perha})s be said for ruminants generally, but from want of e^'idence it is still 
 impossible to form any conclusion in I'egard to whether monoestrum oi' poh'<jestrum is 
 the primiti^'e condition for the Mammalia as a whole. 
 
 The changes through which the sheep's uterus passes during a single dioestrous 
 cycle can be divided into four gioups or periods, which are as follows : — (l) Peiiod of 
 rest ; (2) period of growth and increase of vessels ; (3) period of breaking do^vn of 
 vessels and extravasation of blood ; (4) period of recuperation and pigment forma- 
 tion. These headings indicate the nature of the changes which the uterus undergoes. 
 Bleeding into the uterine cavity and through the external generative opening need 
 not necessarily occur. The extravasated blood left in the uterine stroma forms 
 pigment, which consequently tends to accumulate as the sexual season advances. 
 But the amount of pigment formed depends also on the degree of severity of the 
 prooestrous changes. Where these processes are more severe, there is less extrava- 
 sated blood left behind in the stroma, and so less to take part in pigment 
 formation. Extravasation takes place almost entirely in the more superficial portion 
 of the stroma, the vessels in the deeper part of the uterus remaining intact. There 
 is also some evidence that the severity of the procestrous process tends to diminish 
 with each successive dioestrous cycle in one sexual season. There is never any denu- 
 dation of stroma, and the uterine epithelium disaj)pears only in places where bleeding 
 into the uterine cavity occurs. The pigment formed during the sexual season must 
 
 Digitized by IVIicrosoft® 
 
FORMATION OF THK CORPUS TATTED M TN TIIE SHEEP. 89 
 
 eventually be removed, since I have nevei- found it present in ewes killed during the 
 anoestrous period. 
 
 That the dioestrous cycle of the sheep and the menstrual cycle in the Primates are 
 physiologically homologous is rendered extremely ])rol)a1)le in the liglit of the liicts 
 summarised above, A\4nle it a further comparison bt^ made with otluu' Mammalia, and 
 more particularly the dog and the ferret, the probability becomes converted into a 
 certainty. The procestrmn in the ferret, according to my own researches {an advanced 
 abstract of which is embodied in the present memoir), in regard to its severity is 
 almost intermediate between that of the sheep on the one hand, and the menstruation 
 of the Primates on the other. 
 
 CEstrus in the domestic shee}) very rapidly succeeds the procestrum, so much so 
 that the period of desire seems sometimes to be coincident with the prorestrum. But 
 this is where the procestrous process is less severe. When external bleeding does 
 occur in ruminants, as in bitches and ferrets, there is evidence that it takes place 
 prior to the actual o?strus period, this being known to be the case Avith certain 
 heifers. 
 
 Since the procestrum and menstruation are homologous, and since in at least many 
 ^L\mmalia, coition and the fertilization of the ovum (if not ovulation) occur after 
 the procestrum, it follows that the changes characterising the degeneration and 
 denudation stages of the procestrum and menstruation are not of the nature of an 
 undoing, owing to the absence of a fertilized ovum, of preparations made during the 
 growth period. But the alternative view, that the whole procestrous process (of 
 which the changing of the uterine tissue is a not unimportant part) is in a geneial way 
 an act of preparation, appears to me to be perfectly tenable. On this hypothesis, 
 it does not follow that the process must be directly associated with any particular 
 ovulation. 
 
 With the sheep, ovulation takes place during the sexual season ; it can occur during 
 any oestrus (or possibly occasionally during a procestrum) in the regular sexual season 
 with Scotch Black-faced ewes without the additional stimulus supplied by sexual 
 intercourse. But it would seem that in oestrus periods, which, for one reason or 
 another, exist outside the regular sexual season of the pai-ticular breed, the extra 
 stimulus furnished by coition and the presence of sperms in the uterus, may be 
 necessary before ovulation can be induced. There is evidence also that coition 
 during any oestrus period may hasten forward the ovulation, if it has not already 
 occurred, at that period. 
 
 The development of the corpus luteum (the experimental study of which neces- 
 sarily resulted from the investigation on the periods when the ova ripen) in the sheep 
 was found to resemble that of the mouse and rabbit, as described by Sobotta, in its 
 essential features, the lutein cells being the much-hypertrophied epithelial cells of the 
 undischarged follicle. But the connective tissue element of the corpus luteum in the 
 case of the sheep is supplied not only by ingrowth from the inner theca, but also by 
 
 VOL. CXOVI. B. N 
 
 Digitized by IVIicrosoft® 
 
90 Mi;. F. H. A. MARSHALL ON THE CESTROUS CYCLE AND THE 
 
 the theca externa. The cavity of the corpus luteum becomes filled up by couuective 
 tissue derived similarly from the follicular thecse, and is not formed from leucocytes. 
 But in some of the earlier stages leucocytes are very abundant, wandering inwards to 
 the cavity with the ingrowth of the connective tissue. Later they appear to undergo 
 degeneration, and take no part in the formation of the corpus luteum. 
 
 The atretic follicle can be readily distinguished from the discharged follicle or young- 
 corpus luteum by the absence of any indication of a point of rupture, the degeneration 
 of the epithelial cells, and the character of the connective-tissue ingrowth, which does 
 not occur to an appreciable extent until the epithelium is in an advanced state ot 
 degeneration, or has entirely disappeared. The retention of a degenerate ovum 
 within the cavity is also an almost sure indication of an atretic follicle. 
 
 Scattered red corpuscles may be present in the cavity both in an atretic follicle and 
 also in a young corpus luteum in the case of the sheep. 
 
 I found atresia to set in most commonly among follicles that had reached from 
 about one-half to one-eighth the size of the ripe follicles, but it may occur at any 
 time. When taking place to any considerable extent, it must inevitably raise the 
 barrenness percentage, so that the subject of atresia is not without its practical 
 interest. In Scotch Black-faced sheep (whether pure-bred or half-bred) I found that 
 the number of follicles discharged at a single ovulation only exceeded two in one 
 instance, and that it was far more often one than two. So that evidence points to 
 the conclusion that the number of follicles discharged at an ovulation is, on an 
 average, scarcely in excess of the number of lambs produced at a time. From this it 
 may be inferred that scarcity of twins, or even complete barrenness (whether tem- 
 porary or otherwise) results from failure on the part of the ovary, probably owing to 
 an excess of follicular atresia, whether occurring during the sexual season or at some 
 jDrevious period. 
 
 I have no evidence that oestrus can occur in ewes in the total absence of mature 
 follicles in both ovaries, but I see no reason even in such animals as the sheep, for 
 regarding the connection between ovulation and oestrus to be such an essentially close 
 one as is often maintained. According to all the evidence of sheep-breeders, tem- 
 porary barrenness is not necessarily associated with omission of oestrus (or extension 
 of anoestrum), and I found no relation between the intensity of the oestrous (and 
 procestrous) phenomena and the number of follicles discharged during the particular 
 oestrus. 
 
 Mr. Heape (1900) has laid some stress on the fact that ovulation is not always 
 associated with menstruation in the Primates, and that it may require a stimulus in 
 addition to that supplied by oestrus in the case of the female rabbit ; and in the 
 present paper I have shown that with the bitch ferret while the procestrum and 
 oestrus are distinct periods, the latter may continue for several weeks, during which 
 the animal will permit the occurrence of coition, although the time for ovulation has 
 passed by. 
 
 Digitized by IVIicrosoft® 
 
FORMATION OF THE CORPUS LUTEUM IN THE SHEEl'. <J I 
 
 It would appear tlieii that while the rate of growth and maturation of the follicles 
 are the result of nutriment and general physical condition, hesides being dependent 
 partly upon seasonal changes, the actual rupture of these follicles and the discharge 
 of the ova can, in the sheep and n\any other animals (though certainly not in all), 
 only take place as a result of n special stimulus. In this sense oestrus, and in some 
 instances sexual intercourse, may be said to determine ovulation, but not the periods 
 Avhen ova ripen. But the uterine changes characterising the di(jestrous cycle are 
 themselves in part also the result of nutritive and seasonal influences. So that the 
 phenomena of ovulation and cestrus are connected, though not always necessarily 
 interdependent. 
 
 The relation l)et\\een the de\'elopment of the corpus luteum and the changes 
 occiu'riug in the utei'us during pregnancy is a closer one. According to Born the 
 corpus luteum is of the nature of a gland, the function of which is to secrete into the 
 blood substances which in a general way prepare the uterus for pregnancy.* This 
 view is supported by Fraenkel and Cohn (1901), who have made experiments 
 with rabbits, showing that \^'hen the corpora lutea are destroyed by a galvano- 
 caustic needle, pregnancy cannot continue. Whether Bokn's theory is right or 
 wrong, Fraenkel and Cobn have undoubtedly demonstrated a very intimate con- 
 nection, already knoAvn to be close, between the existence of the corpus luteum and 
 the occurrence of pregnancy in so far as the latter appears to be in some way 
 dependent upon the former. But on the othei' hand, it is a now generally accepted 
 fact that Avhereas the ovarian changes characterising the early development of the 
 corpus luteum ai'e the same, whether or not the discharged ovum is to become 
 fei'tilised (in other woi'ds, that the so-called corpora lutea vera and spuria are by 
 origin identical structures), yet the later changes depend upon the occurrence or non- 
 occurrence of pregnancy. In this sense the further development or degeneration of 
 the coipus luteum may be said to be determined by the state of the uterus. 
 
 From these and similar facts it may be inferred that the functions of ovulation and 
 oestrus do not stand to one another respectively either in the relation of cause to 
 
 * The questions concerning the ovary generally as an organ of internal secretion, and its possible 
 functions apart from the production of ova, have less bearing on the subject of the present memoir. For 
 a recent summary of the work done on these questions, Dixon's paper (1901) may be consulted. It is 
 stated that after the performance of ovariotomy, menstruation genei'ally ceases, the small percentage of 
 cases where it still continues being accounted for on the supposition that both of the ovaries were not 
 completely removed. But ovariotomy, like castration with male animals, has a profound influence over the 
 constitution as a whole, often seriously impairing the health, and not merely preventing the occurrence of 
 menstruation and oestrus. Glass has shown that transplantation of the ovary, after ovariotomy had 
 been previously performed and had induced distressing symptoms and cessation of menstruation, may 
 result in restoration to health and renewal of menstruation (DixON, 1901). The statements of Ferre 
 and Bestion (Dixon, 1901), based on work at present unconfirmed, that injection of ovarian extract may 
 induce genital excitement, more directly concern the subject of this paper, but it is to be noted that 
 unhealthy symptoms, paralysis, &c., were also brought about by the injections. (See also Hbape, 1898, 
 and Halban, 1901. In the latter paper' the effects of ovariotomy with monkeys are described.) 
 
 N 2 
 
 Digitized by IVIicrosoft® 
 
92 MR. F. H. A. MARSHALL ON THE CESTROUS CYCLE AND THE 
 
 effect or conversely, but are connected jDrimarlly, inasmuch as each is dejoendent 
 largely upon the same causes, while a further connection, having regard especially to 
 tlae changes relating to pregnancy, has become established between them, the two 
 sets of functions acting and reacting upon each other ; and that whereas this inter- 
 relation in the majority of the Mammalia is closely maintained, in certain cases, as 
 with the ferret, and more evidently with the Primates, the factors which control 
 the cestrous cycle act to a varying extent independently of those which govern 
 ovulation. 
 
 In conclusion, I wish to thank all those who have assisted me in my investigations. 
 To Professor Ewart, in connection with whose "Biological Farm" at Penycuik, and 
 whose laboratory in the University of Edinburgh the work was carried out, I am 
 especially indebted for the numerous facilities he has placed at my disposal and for 
 the encouragement he has always given. My best thanks are due also to 
 Mr. Walter Heape for giving me the benefit of valuable suggestions and informa- 
 tion, and for the sympathetic interest he has shown, both at the commencement and 
 throughout the progress of my work. To Professor Schafer I am not a little 
 indebted for kindly encouragement and help, more especially with the histological 
 part of the work. Lastly, I take this opportunity of recording my great indebted- 
 ness to Sir Thomas Gibson Carmichael, Bart., for the support he has so 
 generously given. 
 
 References to Literature. 
 
 Aristotle.— 'History of Animals' (Crosswell's translation). Bohn's Library, 
 
 London, 1862. 
 Baer (von). — ' De ovi mammalium et hominis genesi epistola.' Lipsise, 1827. 
 Beard. — 'The Span of Gestation and the Cause of Birth.' Jena, 1897. 
 Belloy.— " Recherches sur I'origine des corps jaune de I'ovaire chez le rat et le 
 
 cochon d'Inde." ' C. R. de 1' Assoc, des Anatomistes,' premiere session, 
 
 Paris, 1899. 
 Beneden (van) and Julix.— " Observations sur la maturation, la fecondation, et 
 
 la segmentation de I'oeuf chez les Cheiropteres." ' Archives de Biologie,' 
 
 vol. 1, 1880. 
 
 Bischoff.— 'Entwickelungsgeschichte des Kanincheneies.' Braunschweig, 1842. 
 Blanford.-" The Fauna of British India." Vol. 1, ' Mammalia,' London, 1891. 
 Bonnet.— " Ueber Melanose der Uterinschleimhaut bei Schafen." 'Deutsche 
 
 Zeitschrift frir Thiermedizin,' vol. 5, 1880. 
 Bonnet.— " Ueber Melanose der Uterinschleimhaut bei brlinstingen und kurze 
 
 Zeit triichtigen Scliafen." ' Deut. Zeitschr. f Thiermedizin,' vol. 7 1882. 
 
 Digitized by IVIicrosoft® 
 
FORMATION OF THE CORPUS LUTEUM IN THE SHEEP. 93 
 
 Bonnet. — " Beitritge zur Embiyoloi^ir der Wiederkiiuei' ,!4('woiiiieu ;uii Scliafe." 
 
 ' Avch. £ Aiiat. 11. Phys., Anat. Al)lli.,' 1884. 
 Bonnet. — Article in Ellen bergev's ' Veigleichende Physiologie der Hanssiluge- 
 
 thiere,' vol. 2, Berlin, 1892. 
 Bonnet and Kolster. — " Benierknngen liber die vergieiohende Histologics der 
 
 Placenta und die Embryotropbit' der Siiugetbiere." ' Verband. d. Anat. 
 
 Ciesellscbaft in Halle,' 1902. 
 Bonnet.— " Beitritge zur Embryologie des Ilundes." 'Anat. Hefte,' vol. 20, 
 
 1902. 
 BuHLBR. — " Entwickelnngsstadien menschlicber Corpora lutea." 'Verband. d. 
 
 Anat. Gesellschaft in Pavia,' 1900. 
 Clark. — " Ursprung, Wachstum iind Ende des Corpus liiteum, &c " ' Arch. f. 
 
 Anat. u. Phys., Anat. Abth.,' 1898 ; American translation in 'Johns Hopkins 
 
 Hospital Keports,' vol. 7, 1899. 
 Dixox. — "The Ovary as an Organ of Internal Secretion." 'The Practitioner,' 
 
 vol. 66, 1901. 
 DoERiXG.— " Beitrag zur Streitfrage liber die Bildung des Corpus luteum." 
 
 'Anat. Anz.,' vol. 16, 1899. 
 Ellexberger. — Article in ' Vergleichende Physiologie der Haussaugethiere,' 
 
 vol. 2, Berlin, 1892. 
 E'wart. — "Variation: Germinal and Environmental." 'Scientific Transactions ol 
 
 the Koyal Dublin Society,' vol. 7, 1901. 
 Fleming. — 'Veterinary Obstetrics,' London, 1878. 
 Flemhing. — " Ueber die Bildung von Kichtungsfiguren in Saugethiereiern beim 
 
 Untergang Graafscher FoUikel." 'Arch. £ Anat. u. Phys., Anat. Abth.,' 
 
 1885. 
 Flower and Lydekker. — ' Mammals, Living and Extinct,' London, 1891. 
 Fraenkel and Cohn. — " Experimentelle Untersuchungen liber den Einfluss des 
 
 Corpus luteum auf die Insertion des Eies." ' Anat. Anz.,' vol. 20, 1901. 
 GiROU. — " Experiences sur la Reproduction des Animaux domestiques." ' Annales 
 
 des Sciences Naturelles,' vol. 11, 1827. 
 Grusdew. — " Versuche liber die Klinstliche Befruchtung von Kanincheneiern." 
 
 ' Arch. £ Anat. u. Phys., Anat. Abth.,' 1896. 
 Halban. — "Ovarium und Menstruation, &c." 'S.B. d. Akad. d. Wissenschaften, 
 
 Wien,' vol. 110, 1901. 
 Hatjsmann.— ' Ueber die Zeugung und Entstehung des wahren weiblichen Eies, 
 
 &c.,' Hanover, 1840. 
 Heape. — " The Menstruation of Semnopithecus entellus." ' Phil. Trans.' B, vol. 185, 
 
 1894. 
 Heape. — " The Menstruation and Ovulation of Macacus rhesus." ' Phil. Trans.,' 
 
 B, vol. 188, 1897. 
 
 Digitized by Microsoft® 
 
94 ME. F. H. A. MAESHALL ON THE OESTEOUS CYCLE AND THE 
 
 Heape. — "The Artificial Insemination of Mammals, &c." ' Proc. Roy. Soc.,' 
 
 vol. 61, 1897. 
 Heape. — " The Menstruation of Monkeys and the Hmiian Female." ' Trans. 
 
 Obstet. Soc.,' vol. 40, 1898. 
 Heapic. — " The Artificial Insemination of Mares." ' The Veterinarian,' 1898. 
 Heape. — " Note on the Fertility of Different Breeds of Sheep, &c." ' Proc. Roy. 
 
 Soc.,' vol. 64, 1899. 
 Heape. — " Abortion, Barrenness, and Fertility in Sheep, &c." ' Jovuiial of the 
 
 Koyal Agricultural Society,' vol. 10, 1899. 
 Heape. — "The 'Sexual Season' of Mammals, and the Relation of the ' Pro- 
 
 testrum' to ' Menstruation.' " ' Q. Jl. Micro. Sci.,' vol. 44, 1900. 
 Henneguy. — " Recherches sur I'Atrcsie des FoUicules de Graaf chez les Mammi- 
 
 feres et quelques autres Vertebres." ' Joui'. de I'Anat. et de la Phys.,' 1894. 
 HoNORE. — " Recherches sur I'Ovaire du Lapin, II. Recherches sur la Forma- 
 tion du Corps jaune." ' Archives de Biologie,' vol 16, 1900. 
 Jan'OSIK. — " Die Atrophie der FoUikel und ein seltsames Verhalten der Eizelle." 
 
 ' Arch. £ Mikr. Anat.,' vol. 48, 1896. 
 Kazzander. — " Uber die Pigmentation der IJterinschleimhaut des Schafes." 
 
 'Arch. £ Mikr. Anat.,' vol 36, 1890. 
 KoELLiKER. — " Uber Corpora lutea atretica bei Siiugetieren." ' Verhand. d. 
 
 Anat. Gesellschaft in Kiel,' 1898. 
 KoLSTER. — " Weitere Beitriige zur Kenntniss der Embryotrophe bei Indeciduateu." 
 
 'Anat. Hefte,' vol. 20, 1902. 
 Keeis. — " Uber Entwickelung und Rlickbilduug des Corpus luteum spurium beim 
 
 Menschen." ' Arch. £ Gymikol,' vol. 58, 1899. 
 Lydekker. — ' Wild Oxen, Sheep, and Goats of all Lands.' London, 1898. 
 Marshall, A. Milnes. — ' Vertebrate Embryology.' London, 1893. 
 Marshall, F. H. A. — "Preliminary Communication on the (Estrous Cycle and the 
 
 Formation of the Corpus Luteum in the Sheep." ' Proc, Roy. Soc.,' vol. 68, 
 
 1901. 
 MiLLAis.— " Two Problems of Reproduction." ' Our Dogs' Publishing Company, 
 
 Manchester, 1895. 
 Nagel.— "Die weibhchen Geschlechtsorgane. " Article in BARDELiiBEN's ' Hand- 
 
 buch der Anatomie des Menschen,' vol. 7, Jena, 1896. 
 Na(!el.— " tJber neuere Arbeiten auf dem Gebiete der Anatomie der weibhchen 
 
 Geschlechtsorgane." Merkel and Bonnet's ' Ergebnisse d. Anat. u. 
 
 Entwickelungs.,' vol 8, 1899. 
 Nathusius.— "Ueber einen auffallenden Racenunterschied in der Triichtigkeits- 
 
 dauer der Schafe." ' Zool. Garten,' Jahrg. 3, 1862. 
 Ott.— " Ueber kiinstliche Befruchtung durch die Peritonealhohle." ' Centralblatt 
 
 £ Gynaek.,' 1882. 
 
 Digitized by IVIicrosoft® 
 
FORMATION OF THI'] CORPUS LUTEIJM TN THE SHEEP. !I5 
 
 Pal.viuno. -"Pel' hi (libutlMla ([uestu^nr sulla (\ss(miz;i del r.nv\>(t liiteo." 'Aiiat. 
 
 Auz.; vol. 18, 1900. 
 FatePvJ^on. — " Ohsevvatioiis on (lorpora [Aitoa." ' Kdiiilmrgh Mnrl. aJid 8ur<)'. 
 
 Journal,' 1S40. 
 Pii.iKVAhSKY.—' Mongolia, the T.ingut Country, a,n(l tlie Solitudes of Noith(u-ii 
 
 Tibet.' (Morgan's Translation), London, 187('>. 
 Rabl. — " Beitrag zur Histologie des Eierstocks, &c." ' Anat. Hefte,' vol. H, 1898. 
 Retterer, — '■ Sur les modifications de la muqueuse uterine .\ I'upoque du rut." 
 
 'C Iv. de la Societe de Bioloyie,' vol. 4, 1892. 
 RoBlNSOX. — " Observations upon the Development of the Common Feiret, Mnsiela 
 
 fcro.r." 'Anat. Anz.,' vol. 8, 1893. 
 Salvi. — " Osservazioni sopra I'accoppiamento dei ( -hirotteri nostrani." ' Atti 
 
 delln Societa Toscana di Scienze Naturali,' vol. 12, 1901. 
 Schottlanber. — " Beitrag zur Kenntniss der Follikelatresie nebst einieen Bemerk- 
 
 ungen liber die unveranderten Follikel in den Eierstocken der Siiugetliiere." 
 
 'Arch. f. Mikr. Anat.,' vol. 37, 1891. 
 ScHOTTLANDER. — " Ueber den Graafschen Follikel, &c." 'Arch, f Mikr. Anat.,' 
 
 vol. 41, 1893. 
 ScHULiN'. — " Zur Morphologie des Ovariums." ' Arch. f. Mikr. Anat.,' vol. 19, 1 88 1 . 
 Scott. — " The Practice of Sheep-Farming." Article in Miles' ' Modern Practical 
 
 Farriery,' London. 
 Shipley. — Article "Nematoda," in 'Cambridge Natural History,' vol. 2, 189G. 
 SoBOTTA. — " Die Befruchtung und Furchung des Eies der Maus." ' Arch. £ Mikr. 
 
 Anat.; vol, 45, 1895. 
 SoBOTTA. — " tjber die Bildung des Corpus luteum bei der Maus." 'Anat. Anz.,' 
 
 vol. 10, 1895. 
 SoBOTTA.— " tJber die Bildung des Corpus luteum bei der Maus." ' Arch. f. Mikr. 
 
 Anat.; vol. 47, 1896. 
 SoBOTTA. — " tJber die Bildung des Corpus luteum beim Kaninchen, &c." ' Anat. 
 
 Hefte; vol. 8, 1897. 
 SoBOTTA. — " Noch-einmal zur Frage der Bildung des Corpus luteiun." 'Arch, f 
 
 Miki'. Anat.; vol. 53, 1898. 
 SoBOTTA. — " tJber die Entstehung des Corpus luteum der Saugethiere." Merkel 
 
 and Bonnet's 'Ergebnisse der Anat. und Entwick.,' vols. 8 and 11, 1899, 1902. 
 Stockel. — " Ueber die cystische Degeneration der Ovarien bei Blasenmole, &c." 
 
 'Sep. Ausdruck aus der Festschrift fiir Fritsch.' 
 Stratz. — ' Der geschlechtsreife Saugethiereierstock.' Haag, 1898. 
 Stricht (van der). — " La ponte ovarique et I'histogenfese du corps jaune." ' Bull. 
 
 de I'Acad. Roy. de Mddecine de Belgique,' 1901. 
 Stricht (van der). — " La rupture du follicule ovarique et I'histogenese du corps 
 
 jaune." ' C. R. de TAssoc. des Anatomistes; Be session, Lyon, 1901. 
 
 Digitized by IVIicrosoft® 
 
96 WE. F. H. A. MARSHALL ON THE (ESTROUS CYCLE AND THE 
 
 Stricht (van der). — ' Une anomalie interessante de formation dii corps jaune.' 
 
 Gand, 1901. 
 Stricht (van der). — " L'atresie ovulaire et I'atrdsie folliculaire du foUicule de 
 
 Graaf, dans I'ovaire de chauve-souris." ' Verhand. d. Anat. Gesell. in 
 
 Bonn,' 1901. 
 Turner. — ' Lectures on the Comparative Anatomy of the Placenta.' Edinburgh, 
 
 1876. 
 Wallace. — 'The Rural Economy and Agriculture of Australia and New Zealand.' 
 
 London, 1891. 
 Wallace. — ' Farm Live Stock of Great Britain.' 3rd Edition. London, 1893. 
 Wallace. — ' Farming Lidustries of Cape Colony.' London, 1896. 
 Weil. — " Beitrage zur Kenntniss der Befi'uchtung und Entwdckelung des Kanin- ■ 
 
 cheneies." ' Wien Mediz. Jahrbuch,' 1873. 
 Wendeler. — Martin's Die Krankheiten der Eierstocke und Nebeneierstocke.' 
 Wiltshire. — " Lectures on the Comparative Physiology of Menstruation." 
 
 ' British Medical Journal,' 1883. 
 
 Description of the Plates. 
 
 PLATE 7. 
 
 Transverse sections thi'ough the uterine cornua, showing portions of the uterine 
 wall, X 220. 
 
 Reference Letteis. 
 
 h. v., blood-vessel; b.v.rup., recently ruptured blood-vessel; c. t., connective tissue 
 of the stroma; ep., epithelium lining uterine cavity ; ep. gl., epithelium of 
 gland ; ex. h. , extra vasated blood corpuscles ; ex. h. p. , ditto in process of 
 transformation into pigment ; /, muscle fibres ; Jae. rup., ruptiu^ed lacuna ; 
 lev., leucocyte; p., pigment; p>.c., pigment cell. 
 Fig. 1. — Section showing a portion of the mucosa of an intercotyledonary area, 
 
 containing a few muscle fibres. (Period I.) 
 Fig. 2. — Section showing a portion of the mucosa of a cotyledon. (Period II.) The 
 stroma nuclei are most thickly distributed in the proximity of the 
 epithelium. 
 Fig. 3. Section showing a portion of the mucosa of a cotyledon. (Period IIL) 
 Fig. 4. Section showing a portion of the mucosa of a cotyledon. (Period III 
 
 advanced stage.) 
 Fig. 5. Section showing a portion of the mucosa of a cotyledon at an early stao-e of 
 
 pigment formation. 
 Fig. 6. Section showing a portion of the mucosa of a cotyledon during a stage of 
 pigment formation. Large quantities of completely- formed pigment 
 lying apparently free in the stroma are also seen. 
 
 Digitized by Microsoft® 
 
FORMATION OF THE CORPUS LUTEUM IN THE SHEEP. 97 
 
 PLATES 8-10. 
 
 Transverse sections through ovaries, sliowing portions of the walls of follicles or 
 developing corpora hitea. 
 
 Reference Letters. 
 
 hi, v., blood-vessel; cav., cavity. {The fluid in the cavity is represented only in 
 fig. 9.) c. <., connective tissue; dep., cup-shaped depression on surface of 
 ovary at point of rupture of follicle (containing islands of epithelial cells, 
 as well as isolated epithelial cells); ep., follicular epithelium or " membrana 
 granulosa " ; ep. c, follicular epithelial cell ; ex. b., extra vasated blood ifus. c., 
 fusiform cell ; leu., leucocyte ; pas., passage leading from depression (repre- 
 senting point of rupture) to internal cavity ; th. int., theca interna ; th. ext., 
 theca externa. 
 
 Plate 8. 
 
 Fig. 7. Section through wall of Graafian follicle X 300. 
 
 Fig. 8. Section through wall of discharged follicle (seven-hour stage) X 300. The 
 ingrowth of connective tissue from the theca interna and the hypertrophy 
 of the follicular epithelial cells have commenced. 
 Fig. 10. Section showing part of wall of discharged follicle (twenty-four-hour stage) 
 
 X 300. The connective-tissue proliferation has become considerable. 
 Fig. 11. Section showing part of wall of discharged follicle (thirty-hour stage) 
 X 100. 
 
 Plate 9. 
 Fig. 9. Section through wall of discharged follicle (sixteen-hour stage) X 300. 
 
 Plate 10. 
 
 Fig. 12. Section showing inner part of wall of discharged follicle (sixty-hour stage) 
 X 100. 
 
 Fig. 13. Microphotograph of transverse section through ovary containing a dis- 
 charged follicle (thirty-hour stage) showing place of rupture and ingrowths 
 
 in the form of strands arising from the theca externa, X 16. 
 
 [Figs. 1-6 were drawn by Mr. J. C. Brodie, of Edinburgh. 
 
 Figs. 7-12 and also tbe microphotograph (fig. 13) were the work of Mr. Richard Muir, of Edinburgh. 
 The clearness and accuracy of Mr. Mum's original drawings are, unfortunately, to some 
 extent lost in the. reproductions. This is seen especially in the absence of outline in the 
 walls of certain of the blood-vessels. — June 2nd.] 
 VOL. CXCVI. — B. O 
 
 Digitized by IVIicrosoft® 
 
Digitized by IVIicrosoft® 
 
AI(irs/inll. 
 
 1' 'i. * 
 
 
 i ••«>■:■'''" ...JSilf,^ 
 
 7^/^/7. TyaK.s: QNol TdbJ'/a/e 7. 
 -big A. .^ -.' , 
 
 
 
 ^*-' 
 
 ■ ^.. ■ vtv ifi. , * ^* V >('^ **■ ^ -H. *»*^ 
 
 .ep.ql. 
 
 
 fcL- 
 
 
 '.fc.-£:. 
 
 »■• "J 
 
 e^. 
 
 fe^---^ 
 
 ^^i^r 
 
 lor rn) 
 
 *^^*W'?. 
 
 ,1 , 
 
 ^ 
 
 6X b 
 
 A 
 
 W*,'^'" ' , iff ep 
 
 ri55.2 
 
 
 J 
 
 
 
 A ^ 
 /?:■ 
 
 
 
 » l- 
 
 "^^ ■*« > 
 
 
 
 ct 
 
 tl 
 
 c.t.- 
 
 bv rap 
 pc... 
 
 rig.5. 
 
 .l.v: 
 
 
 .j-^r 
 
 ct 
 
 - t , - 
 
 
 
 -f'^ » 
 
 r • 
 
 :'bp 
 
 « ' 
 
 ^S* 'v. 
 
 r,^r 
 
 
 
 ex.i 
 
 »;■ 
 
 
 -4> 
 
 ^^-&g^(fby MicrMoft® 
 
 West^ewmaiL chrcriV. 
 
Digitized by IVIicrosoft® 
 
Mar-shaJt. 
 
 Phil 1 
 
 fci/lS. 
 
 \\h>[ ['My Plate Q. 
 
 ^.a. 
 
 Fit;. 7. 
 
 -r^^ ^' ^ OT ^^^ vise 
 
 _ ' ^„rOO D 
 
 m 
 
 
 „0 <-5® 
 
 
 
 :.3^f:?5^ 
 
 
 C3^ C^ ^~^ — ■-^ 
 
 ^j^o V' S^ 
 
 L^rs- 
 
 -,C5 
 
 31 
 
 ':::::> C2> -"e::; 
 
 
 Uu 
 
 6[. 
 
 thxnt. 
 
 
 Fig. 11 
 
 * 1?,S * :, a 
 
 
 
 
 Fig.lO. 
 
 ^lOT^.'^')^-^'^!^ ^f 
 
 fas c 
 
 ■.i<tf?'-V, * 
 
 (^ 
 
 
 6L- 
 
 r ^ /, 
 
 
 .^. 
 
 '-.>*-^ V 
 
 
 ne- 
 
 42E> 
 
 Microsoft®^ ^^^ 
 
 
 
 : rik^ 
 
 
 WeshJ^ewiTiArL chrorric. 
 
Digitized by IVIicrosoft® 
 
Marshn/l 
 
 Ini Tr<ni-s\'Syo\.Yd(iFlote H. 
 
 
 « ^ --f/) 
 0^^ 
 
 
 "^^ * * "T M ^ 
 
 T. <=a« St V S ^ 
 
 
 TC v 
 
 ,« 
 
 ^'*- 
 
 -!*. •*«» 
 
 
 5(5) 
 
 •^B- >w5>- 
 
 
 
 tu 
 
 
 o 
 
 -t/i m^ 
 
 !^5 -on 
 
 -ffl' 
 
 ■•li. 
 
 
 ■i^ 
 
 JOi 
 
 
 " •&. ^- 
 
 C9 , 
 ^"^^ «K£-^«e 'S^"'^<w» '°3bJ!!^^^ ^-aB>'>» 
 
 -^ ._^ ■«* *-^ " -^ 
 
 
 'tS^«* 
 
 
 Wesi:.,Newm^-n cbrcroo 
 
Digitized by IVIicrosoft® 
 
Mars/tail. 
 
 I '1, 1 1 y/Y/^.s,B,V(d. l')6,/^/r/A'10, 
 
 
 -c t. 
 
 Ficil3. 
 
 »;'^>~ 
 
 Digitized by IVIicrosoft® 
 
 "W<;-st.N'^n.TiiaTi cLn-rr.i. 
 
Digitized by IVIicrosoft® 
 
D BY "Harbison and Sons. St. Martin's Lank. 
 Demy 8vo, pp. 427. Price 12s. 6d. 
 THE RECORD OF THE ROYAL SOCIETY. 
 
 Second Edition.' 1901. 
 Containing an Account of the Foundation, History, &c., of the Society, with a Chronological List and 
 Alphabetical Index of the Fellows from the Foundation. 
 
 Beady Shortly. Demy 8vo, pp. 265. Price 5s. 
 THE YEAR-500K 0"F THE ROYAL SOCIETY, 1903. 
 Principal Contents :— -List' of Fellows, Committees, &o.; Statutes and EegulationS; Business of the 
 Society during 1902, including Anniversary Meeting with Pre9i||i]|t*| Address, Couneil's Eeport, 'and 
 Treasurer's Accounts ; Arrangeiiients f or 1903; Obituary Notices of Fellows deceased." 
 
 PHILOSOPHICAL TRANSACTIONS. 
 
 Skmbs a. Cohtaihing Papers of a Mathetnatical or Physical Character. 
 
 Vol. 199 (1902), Price £1 6s. 
 
 Vol. 200 (1903), Price £2 3s., witb 39 Platw. 
 
 Seeies B. Containing P'apers of a Biological Character. 
 
 Vol. 194 (1901), Price £3, with 49 Plates. 
 
 Vol. 195 (1903), Price £1 7s,, with 16 Plates. 
 
 Now' Ready. Price 8s, 
 Eoyal 4to. 112 pages, with Mip, and Illustrations.. '' 
 
 MAGNETIC AND METEOROLOGICAL OBSERVATIONS ,.^,- 
 
 ■ MADE By , _: /', '' ^'-S^'r'*:.' 
 
 THE "SOUTHERN CROSS" ANTARCTIC EXPEpITlON, 1898-1900, '1' ' 
 0nder the direction of C. E. BpBGHGREVlNK. 
 PttbBsfcedby the Royal Society of London, and.sold by HA&BlsOijr and Sons; St. Martin's Lane. 
 Fellows of tlie Royal Society can obtain cppiea gratis, on application. 
 
 Published by Kegan Pat;l,.^ebnch, Tkubnee, And Co;.' 
 A few copies at the reduced price of £1 Is. (published at £3). 
 Eoyal 4to, pp. iv.— 936, cloth. With 51 Lithogriaplji:? and,Photogi;iapM^ Plates; ; 
 A MONOGRAPH OF THE HORl^Y SPONGES. 
 
 By R. von LENDENFELD. 
 A further reduction of price to Fellows' of the ^oy^r Society. 
 
 Published by Kegan Paul, Trench, Tbub^Se,' and C6. 
 Til 1 vol 4to Pn 500 With 6 Chromolithographs of the remarkable Sunsets of 1883 and 
 ■^ ' ^' ■ •■ '' 40 Maps and Diagrams. ,.;, :.. 
 
 THE ERUPTION OF KRAKATOA AND SUBSEQUENT >PHENOMENi. 
 
 • Report of the Krakatoa Committee of the Royal Society. 
 / . " Edited by G. J. SYMONS, F.R.S: , . ; . v ^^:: ' . 
 
 Pj-iee 30.?. To Fellows of the Royal Society and Royal M^,^Qrologieal Soci^yv SOs. ■ 
 
 /Published by Kegan Paul, TpEikiH, Trubner,:^^ 
 Royal 4to, pp: xiv.— 326, cloth. Price 2ls. " 
 
 OBSERVATIONS of' THE mTERNATIQNAL POLAR EXPEDITIQN^ 
 
 '''''■■ J ''"\ ■ ' '• ' ■ 188'2rl'883.', ,_ - "" '■ -'( ■■ ■-' 
 
 ,.,, FORT v-RAK ;■_'..,-, 
 - V With 32 Litl)ogra|h|e, Folding Plates, , 
 Di9kJ0^-£.f .F>>;e. ; v:|!yf^©j);:.ili^ Eoyal ^ociety. ' . ,i ; ' 
 
 ■- ;Cmitinental;AgentsforthepubHcationsofthei^^^ 
 
Sold by Dtjlau and Co.. 37, Soho Square, "W. 
 KECENT PAPERS IN THE 'PHILOSOPHICAL TRANSACTIONS,' 
 
 1902-1903. 
 
 Series A.— ^Mathematical and Physical. " 
 
 A. 333. An ExperimMital Determination of the Variation with Temperature of the Critical Velocity of 
 
 Flow 6f Water in Pipes. By E. 0. COKER, M.^: (Cantab.), D.Sc. (Edin.), Assistant Professor 
 
 of Hy'a Engineering, and S. B. Clement, B.Sc. (McGillX Demonstrator in Civil Engineering, 
 
 both of McGill^^^^^ Mon;treal. Price Is. 
 A.. 334. Gn an Approximate Solution for the Bending of a Beam of Eectangulaf Cross-Section under 
 
 any System of Load, with Special Eeferenfie to Points of Concentrated or Discontinuous 
 ' Loading. By L. N. G. Filon, B.A. (Canl|ib.), M.A., RSo, (Lbnd.), King's College, Cambridge, 
 
 Fellow of University College, LohdoB' and 1851 Exhibition Science . Research Scholar. 
 
 Price 5s. " ''^ . -•, 
 
 A. 335. 'jCJ^'the Vibrations and Stability of a Gravitating. Planet., By J,.H. Jeans, B.A., Isaac, Newtcto.. 
 
 ' Student, and Fellow of Trinity CoUegfe^Canibridge. Price Isi 6A , ; 
 A. 336. On the Formation of Definite Figures by the Deposition Of Dust. By %. J. EUSSELL, Ph.D., 
 
 F.E.S. .Price 2s. 6(?.: ''',-':^ '. 
 
 A. 337. The Spectrum of y Cygni. By^, Norman LqIckter, E.C.B., F.E;S.i and F. E. Baxakdall, 
 
 i A.kC.Sc. IWee2s. 6i "'""^ . ^ 
 
 A. 338. Quaternions and Projectave Geometry. By Professor Charles Jasper Joly.^ Eoyal 
 
 Astronomer of Ireland. Price 5s. 
 A. 339. The, Differential Invariants of a Surface, and their Geometric Sighiidg,jice. By A. E: Forsyte, 
 
 M.A., Sc.D., F.E.S., Sadlerian Prof essor of Pure Mathematics in the University of Cambridge. 
 
 Price 3s. 6(f. 
 
 A. Sio. On the Laws Governing Electric Discharges in Gases at Low Pressures,. By W. R. Carr, B.A., 
 
 Post-,Gradtiate Student, University of Tprontb, Price 2s. ,- 
 
 Series B.—MiologieaK < 
 
 B. 209, The Pharmacology of Pyraconitine and Methylbenzaconine considered in Eelation to their 
 
 Chemical Constijta^ion. By J. Theodore Cash, M.D., JF-Ej^., Eegius Professor of Materia 
 v-.tj^difiaiathe t^i^gity of Aberdeen, and Wyndham E.; Dtjnstan,- M,A., F.E.S., Director 
 
 of the!J^|Entific-and Jjfe|lfeical Department of the Imperial Institute.,' Price Is. 
 B. 210. The ^^^tffe and JDevelofment of the Stem in the Pteridophy%, and Gymnosperms. By 
 
 Edward C. Jeitrey, Ph.D. (Harvard). -iPrjce 3s, ' 
 B. 211. On the Structure of the^Qills oithe Lainellibranehi|i. By W. G. EiDEWOOD, D.Sc, F.L.S., 
 
 Lectiu-er on Biology at the Medical School of St. Mary's Hospital, London. Price 9s. Qd. 
 B. 212. The Development of Echinus esculentus, togethef with some Points in the Development of 
 
 E.milians and E. acdus. By E. W. MacBbide, M A., D.Sc. (LonEt,)i; sometime Fellow of 
 
 St. John's Cblleg^ Cambridge, Strathcona Professor of Zoolggy iia the McGill Universityf 
 
 Montreal. VrkefSs.'i^d. 
 B. 213. Title, Contents, Index^l&c., Vol, 195. Price 6i. 
 B. 214. The Seed-fungus of Lolium temulenium, L., the Darnel. By E. M. Freeman, M,8»> University of 
 
 Minnesota. Price 3s. , 
 
 B, 215. On the Histology|^ Uredo Dispersa, Erikss., aiid the ".MyeopiaBiii" Hypothesis." By 
 
 H. Marshall Ward, Sc.D.,.%E.S., Professor of Botany in the Umver#y of Cambridge. 
 
 Price 3s. 
 
 Cloth cases lor .bin]i^g the ' I^MpgQphi(^r/^^|guHj|H can be obtained of Harrison and Sons, 
 ' St, Martin's Larie. Price Is, 6d. 
 
 Digitized by Microsoft® 
 
Digitized by IVIicrosoft® 
 
Digitized by IVIicrosoft® 
 
Digitized by IVIicrosoft®