On Ligature of the Vas Deferens in the Cat and Researches on the Efferent Ducts of the Testis in Cat, Rat and Mouse

In Rodents the inguinal canal is widely open to the abdominal cavity, in fact the scrotal sac is merely a pouch-like dilatation of the abdominal wall into which the testis with the epididymis fits closely. Contraction of the muscular wall of the scrotal sac forces the testis forwards into the abdomen, and this movement of the testis involves the inversion of the muscular sac, because the cauda epididymis is attached to the posterior end of the sac. The retraction of the testes into the abdomen takes place commonly in the rabbit and rat when the animals are taken up by the hands and the testes descend again when the animal becomes quiet, though sometimes it requires manipulation for several minutes to produce complete descent. It is therefore, as mentioned in my previous paper, possible for the testis in an experiment on the effects of ligature in a Rodent to be partly or completely displaced from the scrotum into the abdominal cavity. In the experiments on the rabbit recorded in my previous paper, in which the vas deferens was reached through an abdominal incision, I was careful to make sure that the testis retained its normal relations to the scrotal sac, and normal spermatogenesis persisted after the ligature. I was desirous, however, of trying the effect of ligature of the vas deferens on some species in which the testis could not be retracted, and in which therefore the possible effect of cryptorchidism was excluded. It is stated in Quain’s Text-book of Human Anatomy that the inguinal canal is actually closed only¹ in the human species, in adaptation to the erect attitude of the body, but it is well known that in many mammals, the testes are not capable of retraction within the abdomen. This seems evident in Carnivora such as the dog and cat, for example, and also in Ungulata such as the horse, bull, etc. The cat appeared to be the most convenient animal for experiment. Dissection showed that in this animal the cavity of the scrotum is connected with the abdominal cavity by a long narrow tube lined with peritoneum, and situated beneath the skin over the ventral surface of the pelvic girdle. There is no muscular layer in the wall of the tube and very little in the wall of the scrotum itself. The inguinal tube contains only the vas deferens and the spermatic blood-vessels and is much too narrow to admit the testis, but at the same time the wall of the tube is lined with peritoneum and not attached by connective tissue to the spermatic cord as it is in the human subject.

Urethane was tried as an anaesthetic for cats but proved very unsuccessful. One cat injected with this drug died after the operation without recovering consciousness, another died during the operation. A successful operation was carried out by using chloroform and ether as anaesthetic. The scrotum on the right side was opened, the vas deferens ligatured in two places and a piece between the ligatures cut out. The animal recovered quickly and lived in good health until it was killed with chloroform 104 days after the operation. The end of the vas next to the operated (right) testis was found to be closed and the spermatic blood-vessels uninjured. Abundant active sperms were obtained from the cauda epididymis of the unoperated side. Sections of the operated testis showed perfectly normal spermatogenesis and when compared with those from the normal left testis showed no structural difference. In both the seminal epithelium exhibited the stages of active spermatogenesis, the sections of the tubules were either in contact or separated by small intervals, and the quantity of interstitial cells, which in the cat are large and conspicuous, was relatively small.

The post-operative period stated by Bouin and Ancel to be sufficient to produce complete disorganisation of the seminal epithelium after ligature and vasectomy in the guinea-pig was 102 days. In this experiment on the cat no injurious effect was visible after the lapse of 104 days. The epididymis of the operated side, as in similar experiments on the rabbit carried out last year, was distended with semen, its diameter being twice as great as that of the normal on the unoperated side. The testis and epididymis of the operated side showed slight congestion of the blood-vessels, having a darker colour than the normal from this cause. The experiment confirms the evidence of the previous experiments on rabbits, and proves that the operation of ligature and resection of the vas deferens does not cause disorganisation of the seminal tubules with cessation of spermatogenesis in the cat within a period of more than three months.

No further experiments were made on cats as it was desired to test the effects of ligature of the vasa efferentia for comparison with those of ligature of the vas deferens, and experiments for this purpose were made on tame albino rats.

In the rabbit and the cat the epididymis is so closely attached to the testis that it is not possible to distinguish the vasa efferentia by inspection of the entire organs, and although it would be possible to pass a ligature round the epididymis or between the epididymis and the testis, such an operation would not occlude the vasa efferentia exclusively. Miss Gertrude van Wagenen, of the University of California, stated in 1924 that in the rat the efferent ducts are sufficiently discrete to permit complete ligation without interfering with the blood supply to the testis. She states that there are from twelve to twenty thin-walled ducts which pass from the head of the testis to the caput epididymis. Examination of the organs in the dead animal for the purpose of identifying the ducts thus described showed that the epididymis is connected with the testis by a delicate membrane about $14$ in. broad which is somewhat thickened where it terminates in a free edge between the extremity of the caput epididymis and the testis. Experiments were made to test the effects of ligation of this thickened membrane, which appeared to be the site of the vasa efferentia as described by Miss van Wagenen, and microscopic investigation was carried out to identify the vasa with certainty. By means of sections small canals were found in the thickened membrane mentioned, but these were destitute of definite epithelium and it was afterwards recognised that they were merely lymph canals.

The relations and course of the true vasa efferentia in the rat were completely followed by continuous series of sections and by clarified preparations of the membrane containing them showing their whole extent from the testis to the epididymis. The number of the vasa efferentia is six, although in some of the sections near their origin from the testis and also at the distal end there are more than six sections of vasa. This is due to the convolutions of the ducts, not to branching. The entire preparation shows very clearly that the vasa unite into one just before reaching the point where they join the caput epididymis, and this single tube forms a lobe which passes to the dorsal surface of the caput and usually projects beyond it in the anterior direction. It then joins the terminal portion of the caput epididymis which is continued backwards into the body of the epididymis. After I had made out these points I received a copy of a valuable memoir on the “Excretory Ducts of the Testicle in Mammalia” by Dr Jacques of the Faculty of Medicine of Strasbourg, to whom my sincere thanks are due. This memoir is illustrated by seven plates and contains the results of a very careful and minute study of the ducts, especially from the cytological point of view. I had already found considerable differences in the relations of the vasa efferentia and the rete testis between the rat and the cat and also between the former and the human species as described in anatomical text-books. In accordance with these differences Benoit divides the Mammalia into two groups :

1. This group includes man and the larger domestic animals, dog, cat, ram, bull, horse. Zoologically these represent the Primates, Carnivora and Ungulata. In these the caput epididymis adheres closely to the testis, but is not connected with it except where the vasa efferentia arise from it. Each vas forms a cone with its apex at the surface of the testis (conus vasculosus) and the caput epididymis is formed almost entirely by the vasa efferentia.

2. This group consists of the mouse, rat, guinea-pig and rabbit. In these the caput epididymis is not directly attached to the testis, but connected with it by a tract of 5 to 10 mm. length, formed by the vasa efferentia. The caput is composed essentially of the epididymal canal, except in the guinea-pig where it is formed partly by the vasa efferentia. This group zoologically represents the Rodentia.

I will now describe the structure of the vasa efferentia in the rat and of the parts immediately connected with them in the testis and epididymis, pointing out where my observations agree with, or differ from, the descriptions of Dr Benoit.

The relations of the various parts are represented diagrammaticallyin Fig. 1, p. 15. The scrotum is in communication with the abdominal cavity by a wide opening. A thin membrane passes from the peritoneal lining of the scrotum to the vas deferens and thence to the epididymis. The base of the caput and body of the epididymis towards the testis is a smooth concave surface, and the dorsal edge of this is connected to the testis by a thin membrane, which curves round on the surface of the testis and ends in a free edge (M.M. in figure). The margin is slightly thickened, and is termed in the description of my experiments the marginal membrane. From the anterior end of the epididymis a conical mass of fatty tissue projects into the abdomen, the fat-body, F.B. The spermatic vessels, much convoluted, form a vascular cord which passes dorsal to the epididymis and joins the testis on its outer side: one arterial branch, with an undulating course along the inner edge of the testis, is conspicuous. The vasa efferentia leave the testis in the angle between the vascular cord and the surface of the testis and pass forwards and inwards, dorsal to the epididymis, and after forming a bulbous enlargement are continued into the anterior lobe of the caput, which reverses direction at the free extremity of the latter and passes back into the body of the epididymis, then forms the cauda epididymis and is finally continued into the vas deferens.

These have a rather small lumen and thin walls (Pl. I, Fig. 2). They are separated from each other by loose connective tissue containing many fat cells, and the whole bundle is enclosed by a layer of denser fibrous tissue. The epithelium is columnar and has under a low magnifying power a characteristic appearance due to the presence of two layers of nuclei, a basal layer of numerous crowded nuclei and a distal layer, nearer to the lumen, of fewer nuclei at a greater distance from each other. These belong to two kinds of cells described in detail by Benoit (loc. cit. p. 24, PL II, fig. 24). I can only accept Benoit’s description, my own observations not having been carried out so minutely. The more numerous cells with the basal nuclei are secretory cells, terminating distally in a bordure en brosse or striated border, composed of short filaments. Benoit does not regard these filaments as cilia and makes no mention of movement in them. The distal nuclei belong to non-secretory cells each of which terminates in a tuft of long cilia, very conspicuous in my preparations. These ciliated cells are narrowed at their basal ends, while the secretory cells are somewhat broader towards the base than distally.

The rete testis in the rat is situated at the surface of the testis at about one-third of the length of the organ from its anterior end, and adjacent to the attachment of the membrane connecting the testis with the epididymis. Into the base of this membrane, which is thickened with adipose tissue, the connective tissue capsule of the testis projects as a ridge which contains the reticular cavities of the rete and their continuation into the vasa efferentia (Pl. I, Fig. 1). Internally, posterior to the origin of the vasa, tubular branches of the rete extend to the inner surface of the capsule and spread out laterally where they take origin from the seminal tubules. These tubes of the rete correspond to those which have been called vasa recta in human anatomy.

The chambers of the rete have a very characteristic appearance, in strong contrast to the seminal tubules on the one hand and to the vasa efferentia on the other. The cavities are large and the epithelium very thin, composed of a single layer of small short cells at first appearing non-ciliated, but most of the cells have a single cilium in the middle of the free surface. The vasa efferentia arise simply as branches of the rete which pass outwards and anteriorly and of which the epithelium, without any gradual transition, changes to that above described in the vasa. Sections of these branches often show short lengths of the high, ciliated epithelium of the vasa alternating with tracts of the thin epithelium of the rete.

In the other direction can be traced the junctions between the seminal tubules and the inner branches of the rete. The seminal tubules in the neighbourhood of the rete run longitudinally, and the transition from these tubules to the branches of the rete can be traced in a series of transverse sections of the testis in this region. It is a curious fact that there is no direct, open, communication between the central cavities of the seminal tubules and the empty cavities of the rete. In a young rat in which spermatogenesis was only beginning the transition in a single tubule is as follows. In a section of a tubule showing the beginning of the transition the spermatic epithelium on one side of the tubule is still normal, consisting of several layers of cells, cells of Sertoli with small, slightly stained nuclei at the base, and over these several layers of well-defined cells, some with very large nuclei; these are the spermatogonia in stages of subdivision. On the other side of the tubule are a number of rather small nuclei irregularly scattered in a reticulum of fibrils with no definite cell-boundaries. According to Benoit and Regaud, whom he quotes, these cells (in testis of bull and guinea-pig) are the cells of Sertoli, the seminal elements having disappeared, but in my preparations, although the cells in question appear to be continuous with the Sertoli layer, they are no longer arranged as an epithelium but irregularly scattered through a thick tissue. The tube still has a small lumen. Nearer to the rete the tubule becomes slightly smaller, spermatic cells have disappeared altogether and the other tissue with scattered nuclei extends all round the section with still a small lumen in the centre. The tubule then becomes reduced in thickness to half that of the original tubule, and the lumen disappears altogether. In the next stage the deepest layer of nuclei assumes an epithelial arrangement and becomes spread out over the membrane surrounding the tubule, while a lumen appears between them and the rest of the reticular tissue which occupies the middle of the tubule. The sections of the tubule are now slightly oblique because it is slanting outwards towards the rete, and there is a plug of reticular tissue, in which the nuclei are now more radially arranged, projecting into the cavity of a vas rectum. A little beyond this the reticular tissue disappears altogether, the lumen of the vas rectum is wide and empty, is entirely surrounded by the very narrow epithelium, and opens into the spacious central cavity of the rete testis.

As already stated the number of vasa efferentia in the rat is six, and they are contained in the membrane connecting the dorsal edge of the base of the epididymis with the testis. The bundle of vasa arises from the surface of the testis in the angle between this surface and the vascular cord containing the convoluted spermatic artery and vein. The vascular cord passes directly forwards dorsal to the epididymis and enclosed in the fat-body which extends in a conical shape from the caput epididymis forwards. The cord then bends almost at right-angles to pass towards the middle line of the dorsal wall of the abdominal cavity. The bundle of vasa efferentia passes obliquely towards the mid-line of the body till it reaches the caput epididymis, which it passes on the dorsal side. As shown in the diagram, p. 15, there is room to pass a ligature round the bundle of vasa without injuring the vascular cord or any other parts. The bundle of vasa is enlarged distally in consequence of the convolution of the tubes, and this rounded enlargement projects anteriorly beyond the surface of the caput. After a slight constriction the bundle of vasa then joins an anterior lobe of the caput and this at the free extremity of the latter passes into the posterior lobe which has a concave surface adjacent to the convex surface of the testis, and which is continued posteriorly into the body of the epididymis. As previously mentioned the vasa efferentia unite into a single tube. This occupies most of the enlargement above mentioned and then enters the anterior lobe of the caput. It is lined by epithelium of the same character as the vasa efferentia and it is succeeded by the initial segment of the epididymal tube which is characterised by its much greater thickness and by its very high epithelium (Pl. I, Fig. 3). The cells are all furnished with tufts of cilia and distal nuclei are less numerous than in the vasa efferentia. The epithelium is twice to three times as high as that of the rest of the epididymal tube. Another peculiarity of the initial segment is that the cavities of the sections of the tube contain only few sperms or none at all. The probable explanation of this is that suggested by Benoit, namely, that the lumen of the initial segment is much narrower than that of the epididymal tube, and that therefore the liquid secretion flows through it more quickly than through the latter, carrying the sperms into the wider part of the epididymal tube, where they accumulate.

The main part of the epididymal tube has a much narrower epithelium, which is ciliated; distal nuclei are rare. The transition from the initial segment to the main tube occurs at the free apex of the caput where the anterior lobe is continued into the posterior lobe in which the tube reverses its direction and passes back to the body of the epididymis, which runs down the outer side of the testis. In the cauda epididymis the mass of convolutions becomes wider and continues to the posterior extremity, where it reverses its direction as at the apex of the caput and forms an inner lobe which leads to the vas deferens. The tube in this inner lobe reaches its maximum thickness and largest lumen, and here it is surrounded by a thin coat of unstriated muscle-fibres, which are difficult to recognise in other parts of the tube, though Benoit states that they can be demonstrated by special staining in all parts of the tube, and even on the vasa efferentia.

The following are measurements of the median dimensions of the various parts of the efferent system of the testis above described, slight variations above and below these measurements being of course frequent:

The structure and relations of the efferent ducts in the Mouse are very similar to those above described in the Rat. As in the latter the epididymis is connected with the testis by a membrane of some breadth and is not directly attached to the testis. The parts being small the whole of the organs can be dehydrated and cleared in xylol or oil of cloves and examined under a low magnifying power. The bundle of vasa efferentia joins the caput epididymis at a point nearer to its free extremity than in the rat. The seminal tubules are connected with the branches of the rete in the same way, each tubule becoming contracted and filled with a plug of cellular tissue from which a conical process projects into the lumen of the vas rectum. In the specimen which I examined there were three vasa efferentia, but according to Benoit there may be four or five. As in the rat the vasa unite into a single tube which is continued into the initial segment of the epididymis.

My observations on the efferent structures in the Cat agree in general with those of Benoit, but differ in some points, and I have made no attempt to follow out completely, as he has, the course of the vasa efferentia. My intention here is to describe the most important features seen in a series of sections transverse to the axis of the testis, and including both testis and epididymis.

The main part of the rete testis is axial and extends far back towards the posterior end of the testis. Seminal tubules pass towards the rete in radial directions and end by becoming connected with radial branches of the rete. The connection is of the same kind as that described above in the rat, there is no open communication between the cavity of a seminal tubule and that of the rete. The cone of cellular tissue blocking the tubule and projecting into the cavity of the rete is larger and more conspicuous than in the rat, and as the connecting parts of seminal tube and rete are at right angles to the axis of the testis they are in transverse sections of the testis often cut longitudinally (Pl. I, Fig. 4), and sperms are seen imbedded in the cone of tissue as well as beyond it in the tube of the rete. The cells of the cone are elongated and converge obliquely to the free apex of the cone where there is a group of nuclei which are smaller than those in the main part of the cone. The appearance suggests that the cells are not all cells of Sertoli, but actually represent the seminal cells or spermatogonia, which are not here undergoing the mitotic divisions, of spermatogenesis, but are dividing amitotically.

Since artificial cryptorchidism and also ligature of the vasa efferentia cause degeneration of the spermatic part of the seminal epithelium, leaving the cells lining the basal membrane, i.e. as usually stated, the cells of Sertoli, uninjured, this result would appear to afford a criterion whereby to decide whether the terminal cone is really composed of cells of Sertoli or of gametic cells. I therefore examined my serial sections from testes of a rat in which the vasa efferentia had been ligatured 14 days before it was killed. I found that in these sections the ends of the seminal tubes were in open communication with the branches of the rete, the terminal cone having disappeared. This fact supports the conclusion that the terminal cone is modified spermatic tissue and that the cells of Sertoli are continued as the epithelium of the rete.

There are three possible ways in which the sperms may be carried through the terminal cone of the seminal tubules: (1) they may penetrate and pass through the soft plasm of the cells by their own activity, (2) they may be carried through by movement of a secretion produced by the cells, (3) the cells themselves may be dividing without mitosis and breaking up and dissolving at the end of the plug carrying the sperms with them. The appearance of the mass of small nuclei at the end of the cone is certainly in favour of the third hypothesis. In many cases, however, a little mass of crowded sperms is seen escaping from the cellular cone at its side, near its base, and this would support suggestion (1) or (2).

The central cavities of the rete are large and irregular, lined with small, short, almost flat cells with deeply staining nuclei, as in the rat. Farther forwards these cavities communicate through a longitudinal radial plate of interstitial tissue with a superficial portion of the rete spreading out at the inner surface of the testis capsule and triangular in section. This is situated on the inner side of the vascular plexus formed by the spermatic arteries and veins where they enter or leave the testis, and near the attachment of the body of the epididymis. Traced still farther forward the superficial rete is found communicating with a number of very large cavities situated in the outer part of the testis capsule and in the connective tissue extending from the capsule to the epididymal structures (Pl. I, Fig. 5). These cavities, where they communicate with the rete, are lined with the same kind of epithelium as the latter, on the outer side they are lined with a higher, columnar, ciliated epithelium like that of the vasa efferentia. This group of large cavities is not present in the rat, but it corresponds to the initial portions of the vasa efferentia where they originate from the rete. In the rat the initial parts of the vasa are not much larger than the other parts, in the cat the difference is extreme. The large cavities above described when traced forward diminish much in size and are continued into a mass of very small convoluted tubes, the distal parts of the vasa. The largest of the initial cavities measure 320 μ in diameter, while the corresponding measurement of the distal parts is only 75 μ, that is less than a quarter of the diameter of the initial part. Some of the sections of the distal parts are smaller than this, only. 40 μ in diameter, and in these the lumen is extremely small, the cilia actually meeting in the centre. This extreme difference in thickness between the initial parts of the vasa and the distal parts in the cat is not mentioned by Benoit. He states that the vasa open in succession into the initial part of the epididymal canal. I have not attempted to trace out these openings, but have followed the mass of convoluted vasa forward till they come into contact with the convolutions of the initial segment of the canal, distinguished in the cat, as in rat and mouse, by the very high ciliated epithelium and the scarcity or complete absence of sperms in the lumen. This segment lies in an anterior lobe of the caput, as in the rat, and is continued at the free end of the caput into the main epididymal tube which reverses its direction and passes back to the body of the epididymis. As usual, the cavity of the epididymal tube, except the initial segment, is, in the normal mature cat, filled with sperms.

In the first of these the rat was killed three weeks after the operation and sections of the testis showed complete disorganisation of the seminal epithelium in the tubules. Subsequent experiments showed that this was not a constant result. The following is a list of these experiments:

Marginal membrane in rat ligatured on right side only:

Marginal membrane ligatured on left side. Right testis detached from scrotum and fixed to abdominal wall by ligature passing through gubernaculum :

Of the first six experiments, leaving out of consideration that in which the other testis was removed from the scrotum to the abdomen because in that the ligature was found to be loose, there was complete disorganisation of the seminal epithelium in two, the first and fifth; normal spermatogenesis in one, the fourth; and slight traces of disorganisation in a few tubules in three. Careful examination by sections and cleared preparations showed nothing in the marginal membrane but lymphatic channels and small blood-vessels. It is probable that the injurious effects of ligature of this membrane when they occur are due to varying degrees of injury to the tubules of the apex of the caput epididymis to which the membrane is attached.

The following is a list of the experiments in which the true vasa efferentia were ligatured :

Vasa efferentia ligatured on left side, marginal membrane on the right:

Vasa efferentia ligatured on right side only:

These experiments show that closure of the lumen of the vasa efferentia causes complete disorganisation of the seminal epithelium in seven days, or even in six days, assuming that the evidence of the first experiment is valid. In the other three the left testis served as control, and showed perfectly normal spermatogenesis. As there was no interference with the blood circulation in the testis, the effect must be due to increase of pressure within the tubules. It follows therefore that when the vas deferens only is ligatured, the absence of injurious effect is due to the fact that the great space contained in the long coiled tube of the epididymis prevents this increase of pressure in the seminal tubules, the epididymis acts as a reservoir for the semen and becomes greatly distended in consequence. The conclusion is that when the distension and pressure increase to a certain point within the epididymis, absorption of semen is increased, an equilibrium is reached, and no further rise of pressure occurs.

In the testis of Experiment 2 on ligation of the vasa efferentia, killed 14 days after operation, the seminal epithelium is very much reduced. The majority of the tubules show a single layer of nuclei next to the basement membrane covered by an irregular layer of flocculent debris, and a large empty cavity in the middle of the tubules. In some of the tubules there are a few giant nuclei in addition to the parietal layer. Some of the tubules again have no lumen but are completely filled with the flocculent material derived from the disorganised epithelium.

In both normal and operated testes the tubules vary considerably in thickness. In the normal testis the maximum transverse diameter of a tubule is about 350 μ. As a result of the operation the thickness of the tubules is much reduced, in the case of the second experiment the diameter is from 175 to 200 μ. The testis from Experiment 3 in the same list, in which the post-operative period was also 14 days, shows a similar histological condition in the tubules. The giant nuclei are very large and spherical, and contain within them a large number of small spherical bodies usually peripheral in position, each of which appears to consist of one hemisphere which is transparent and highly refractive and another which stains deeply in haematoxylin. The most remarkable feature in the sections of this testis is an enormous cavity at the anterior end which is surrounded by a thick cellular layer composed of many small cells, not arranged regularly as in an epithelium. The cavity is filled with spermatozoa. The inner larger portion of the cellular wall has the cells arranged perpendicularly to the surface, the outer portion is thinner and its cells are parallel to the surface. A basement membrane is not everywhere distinct, in places the cellular wall is continued into the interstitial tissue. I have found no opening from this sac either into the rete or into the seminal tubules. The latter can be traced into a terminal portion where the lumen disappears and the tubule is completely filled with a mass of sperms surrounded by the layer containing nuclei. The end of the tubule finally merges into the cellular tissue forming the wall of the large sac. The large sac seems to be formed by the union of many small sacs arising from the ends of tubules in this way. The internal pressure seems to cause the spermatozoa to collect at their blind ends and then to pass out into the interstitial tissue whose cells multiply and surround the mass of semen with a cellular wall.

In the testis from Experiment 4 in which the post-operative period was only 7 days, the seminal epithelium is, as would be expected, much less disorganised and reduced. Most of the tubules in section have a central lumen, in others the cavity is filled with fragments of cells among which are the remains of spermatozoa. In none of the tubules are normal stages of spermatogenesis to be seen (PL I, Fig. 6). The epithelial layer is still thick, it contains numerous nuclei of spermatogonia. These nuclei have lost their regular arrangement, their chromatin is in the convoluted prophase and they are of various sizes. All stages of transition can be seen, from the larger of these nuclei to large giant nuclei of the structure already described. In a few cases the epithelium is reduced to a thin parietal layer with little more than a single layer of nuclei, in fact nearly to the condition described in the testis after a post-operative period of 14 days. The reduction in transverse diameter of the seminal tubules is naturally less in the testis prepared after 7 days than in that in which degeneration proceeded for 14 days. In the testis from Experiment 3 the maximum diameter was 225 μ while in that from Experiment 4 the maximum was 270 μ. In the latter case the range of variation was very great, in some tubules the diameter being no more than 170 μ.

In the rat the interstitial tissue seems to be present in much smaller quantity than in the cat or rabbit, but the difference is chiefly due to the small size of the special interstitial cells in the rat compared with those in the other two cases. Of the three animals the cat has the largest interstitial cells. In the two rats killed 14 days after ligature of the vasa efferentia the amount of interstitial tissue between the tubules was considerably increased but the total amount may have been increased only relatively as a necessary consequence of the reduction in thick ness of the seminal tubules. In the rat killed 7 days after ligature of the vasa efferentia the relative increase of interstitial tissue was scarcely less apparent than m those which had lived 14 days after the operation.

I have now to compare briefly the condition of the seminal epithelium in Experiment 1 of the vasa efferentia series, and in those in which disorganisation of the epithelium followed ligature of the “marginal membrane.” In Experiment 1, in which the vasa efferentia were ligatured on the left side and the marginal membrane on the right, both testes immediately after death were found to be small and flaccid, but the right larger and firmer than the left. The vas deferens did not show the white thread of contained semen which is seen in the normal mature animal. Microscopic examination of the contents of the vas deferens showed a few scattered spermatozoa, and. still fewer could be obtained from the cauda epididymis, but in both cases they were motionless. Rather more numerous spermatozoa were obtained from the left vas deferens and these were moving, but the movement was feeble. Stained sections from both testes compared with those from Experiment 4 of the same series, from a testis 7 days after operation, present a marked contrast. In Experiment 1 the epithelium in the majority of the tubules is thick and consists of several layers of cells with large nuclei.; it has a regular appearance which does not suggest disorganisation, and resembles closely the structure seen in a normal immature rat. On the other hand, in a few of the tubules the inner cells are loose and in some there are giant nuclei. These signs of disorganisation occur in both testes but more frequently in the left, in which the vasa efferentia were tied, than in the right, in which the ligature was applied to the marginal membrane. I conclude that this experiment was carried out on a young immature rat in which spermatogenesis had scarcely begun and that ligature has had little effect in the short post-operative period of 6 days.

In the sections of testes showing disorganisation of seminal epithelium as a result of ligature of the margin of the connecting membrane, the histological condition is somewhat different in the two cases Nos. 1 and 5 in the list of operations of this kind. In the first, in which the post-operative period was 21 days, the condition resembles that already described as seen after ligature of the vasa efferentia, but in No. 5 of the list, with a post-operative period of 17 days, the tubules are almost or quite full of disorganised material containing numbers of giant nuclei of the structure already described.

In Rodents the inguinal canal is represented by a wide opening and the testis can be retracted into the abdomen. In the cat and other Carnivora the inguinal canal is open but narrow, and retraction is not possible. The operation of vasectomy was performed on a cat which was killed 104 days after the operation. Anatomical and microscopic examination of the testis of the operated side showed perfectly normal spermatogenesis.

In the rat and mouse the epididymis is connected with the testis by a membrane of some width in which the vasa efferentia are contained, and a ligature can be passed round them. In-the cat the epididymis is closely attached to the testis and the vasa cannot be separately ligatured.

The cavity of the seminal tubules has no open communication with that of the branches of the rete testis: the seminal epithelium ends in a solid cone of cellular tissue and the spermatozoa pass through this cone into the cavity of the rete.

In one set of experiments a ligature was passed round the free margin of the epididymo-testicular membrane, and in another set round the vasa efferentia in the rat. The results showed that ligature of the vasa efferentia is always followed by disorganisation of the seminal epithelium, and cessation of spermatogenesis. After the ligature of the marginal membrane in the majority of cases normal spermatogenesis continued, in a minority the seminal epithelium was disorganised.