In the seminiferous tubules and interstitium of the viper there occur cyclical changes basically similar to those in birds and the pike. These involve the metamorphosis of unshed germinal material into cholesterol-positive lipids, the rehabilitation of the interstitium after the seasonal depletion of its cytoplasmic lipids and cholesterol shortly after the spring emergence from hibernation, and the completion of spermatogenesis after basking in the sun. The viper differs from birds in that its tubule lipids contain much less cholesterol; also, the subsequent spermatogenesis begins almost immediately (as in poikilothermous pike and frog), and proceeds to the spermatid stage in the presence of such lipids. Further, although in some individuals new Leydig cells rapidly mature, this resurgence is delayed in others. Thus, although the succeeding spermatogenesis always begins in still-lipoidal tubules, it may do so in the presence of a largely non-lipoidal interstitium. Mitoses were observed in mature Leydig cells. Fibroblasts occasionally alter to take on an apparently glandular function.

By July most animals had regained a heavy lipoidal and cholesterol-positive interstitium, lost most of their tubule lipids, and possessed either secondary spermatocytes or, sometimes, spermatids. This internal activity is probably partly responsible for autumnal sexual behaviour. Factors additional to temperature control spermatogenesis, for only a few spermatozoa are formed by the time of hibernation in October. No pronounced retrogressive changes occur during hibernation: the cycle merely halts, and begins without further delay on emergence. Thus spermatogenesis may be at its peak by the end of March.

The expanded, glandular ‘sexual segment’ of the male renal tubule also exhibits cyclical changes involving the production and expenditure of cholesterol-positive lipids, but such changes take place also in narrow parts of the tubule and, further, in the kidney of the female.

With two plates (figs, 1 and 2)

AT the conclusion of spermatogenesis in birds, the seminiferous tubules ‘L A undergo a metamorphosis involving the sudden breakdown of unused &M’minal material and the genesis of a large quantity of cholesterol-positive lipids. At about the same time a new generation of Leydig cells arise in the exhausted interstitium (Marshall, 1949). The tubule lipids and contained cholesterol, the function of which are unknown, disappear at rates varying between species and with the environmental conditions to which they are subjected (Marshall, 1955; Serventy and Marshall, 1956). Although a few primary spermatocytes may arise in the presence of such lipids, the last remaining traces of fat always disappear at the true onset of the succeeding spermatogenesis.

Such a complete post-nuptial tubule metamorphosis has not been shown in mammals except after hypophysectomy (Coombs and Marshall, 1956), but it is a regular part of the cycle of certain teleost fishes (Marshall and Lofts, 1956). The present study was designed to determine whether comparable events occur naturally in the tubules and Leydig cells of Reptilia and if, in addition, cyclical lipid changes take place in the supposedly sexual segment of the kidney.

Vipera b. berus (L.), the common, northern or European viper or adder, is widely though discontinuously distributed throughout Europe, including Britain. It extends also across Asia as far as China and Sakhalien (Smith, 1951). In Britain hibernation generally begins in October and ends in February or early in March, when groups of emergent vipers are sometimes found coiled together in the sun. Later, males face each other and perform a harmless writhing ‘dance’ which may be of territorial significance. These dances often occur near a female. Afterwards the male displays directly to a female and copulation occurs in April or May. Like many species of birds (Morley, 1943; Marshall, 1951), the viper undergoes autumn sexual activity as well.

The present study is based on 55 adult males (minimum length 480 mm), 3 juvenile males (average length 369 mm), and odd adult females collected during two seasons at Carradale, Scotland (lat. 55.36 N), and Doncaster, England (lat. 53.31 N), in addition to 4 adult males and 5 adult females experimentally hibernated in London during the winter. The hibernated animals were enclosed on 20 September in a dark but ventilated vivarium measuring 5 by 3 by 2 feet and left in a semi-open situation exposed to normal temperature fluctuations but not to rain, wind, or direct light. They were killed on 30 January.

The vipers were caught by means of artery forceps and killed by decapitation. Fragments of testis, kidney, fat body, and epididymis were removed and pieces of each tissue fixed in formaldehyde-calcium (FC) or Bouin’s fluid. The FC material was embedded in gelatine, sectioned at 10 μ on the freezing microtome, coloured with Scharlach R for the demonstration of lipids, and counterstained with Ehrlich’s haematoxylin. Sections were submitted also to the Schultz test for cholesterol, with a section of the adrenal as a control. (Subperitoneal depot fat was never positive to the Schultz test.) The Bouin material was embedded in paraffin wax, sectioned at 7 μ, and stained with iron haematoxylin for the examination of spermatogenetic stages. All measurements in microns refer to paraffin sections.

The testis falls into the typical vertebrate pattern of seminiferous tubules and prominent glandular interstitial Leydig cells (fig. 1, A, B). The presumed sexual segment of the kidney (Gampert, 1866; Regaud and Policard, 1903) may constitute up to one-half of the total organ in some males. Viewed in cross-section, the male kidney is seen to contain two distinct tubular components. Each appears to represent a different part of the same renal tubule: the narrow (about 30 μ.) renal element is proximal to the glomerulus ; the wider (up to 180 μ), obviously glandular, ‘sexual’ structure (fig. 2, c) arises sub-terminally in sharp transformation. This segment is composed of columnar cells (fig. 2, D) in which, apparently, secretion granules have been demonstrated. It is said to be most active during the spring maturation of spermatozoa, when supposed secretory products are discharged into the lumen (Volsoe, 1944). A profuse capillary system ramifies in the interstices between the tubules.

FIG. 1.

(plate). A, 31 March. Testis approaching height of spermatogenesis. Huge numbers of spermatozoa (some indicated by top ring) are generally masked by tubule lipids that are weakly positive for cholesterol. Interstitial Leydig cells (bottom ring) have been depleted of lipids and cholesterol (compare B). (FC, 10 μ, Scharlach R.)

B, 16 April. Testis showing heavy tubule steatogenesis and pronounced interstitial rehabilitation at the conclusion of spermatogenesis. (Technique as in A.)

c, 18 May. Testis showing cessation of spermatogenetic activity, despite interstitial rehabilitation. A few spermatozoa remain unshed. The clear tubule lumina and sub-peripheral vacuoles held lipids in the living reptile. (Bouin, 7 μ, iron haematoxylin.)

D, 18 May. Testis showing tubule clearance proceeding in an animal with largely exhausted interstitium (ringed). (Technique as in A.)

E, 31 May. Testis showing renewal of spermatogenesis with the production of secondary spermatocytes. An interstitial mitosis is ringed. (Technique as in c.)

F, 24 July. Testis with tubules almost cleared of post-nuptial lipids and now full of developing germ-cells alongside a fully regenerated interstitium. (Technique as in A.)

FIG. 1.

(plate). A, 31 March. Testis approaching height of spermatogenesis. Huge numbers of spermatozoa (some indicated by top ring) are generally masked by tubule lipids that are weakly positive for cholesterol. Interstitial Leydig cells (bottom ring) have been depleted of lipids and cholesterol (compare B). (FC, 10 μ, Scharlach R.)

B, 16 April. Testis showing heavy tubule steatogenesis and pronounced interstitial rehabilitation at the conclusion of spermatogenesis. (Technique as in A.)

c, 18 May. Testis showing cessation of spermatogenetic activity, despite interstitial rehabilitation. A few spermatozoa remain unshed. The clear tubule lumina and sub-peripheral vacuoles held lipids in the living reptile. (Bouin, 7 μ, iron haematoxylin.)

D, 18 May. Testis showing tubule clearance proceeding in an animal with largely exhausted interstitium (ringed). (Technique as in A.)

E, 31 May. Testis showing renewal of spermatogenesis with the production of secondary spermatocytes. An interstitial mitosis is ringed. (Technique as in c.)

F, 24 July. Testis with tubules almost cleared of post-nuptial lipids and now full of developing germ-cells alongside a fully regenerated interstitium. (Technique as in A.)

FIG. 2.

(plate), A, 3 January (in experimental hibernation). Testis with all germinal stages up to spermatids. Mitotic activity, however, has stopped (compare fig. 1, E). (Technique as in fig. i, c.)

b, juvenile (365 mm long) 30 June. Testis with primary spermatocytes and a lipoidal, cholesterol-positive interstitium. There has been no advanced spermatogenesis and therefore no metamorphosis (compare fig. 1, A, B, D). (Technique as in fig. 1, A.)

c, male kidney (18 May), showing cholesterol-positive lipids in basal epithelium (arrowed) and aggregations of the same material in the tubule lumen. (Technique as in fig. 1, A.)

D, male kidney (18 May), showing depletion of lipid elements. (Technique as in fig. 1, A.)

E, male kidney (18 September), with lipid-free ‘sexual-segment’, but with apparent renal dements (arrowed) pronouncedly lipoidal and cholesterol-positive. (Technique as in fig. 1, A.)

F, female kidney (20 June), showing an appreciable steatogenesis in the larger elements which probably correspond with the male ‘sexual segment’ shown in fig. 2, c, D, E. (Technique as in fig. 1, A.)

FIG. 2.

(plate), A, 3 January (in experimental hibernation). Testis with all germinal stages up to spermatids. Mitotic activity, however, has stopped (compare fig. 1, E). (Technique as in fig. i, c.)

b, juvenile (365 mm long) 30 June. Testis with primary spermatocytes and a lipoidal, cholesterol-positive interstitium. There has been no advanced spermatogenesis and therefore no metamorphosis (compare fig. 1, A, B, D). (Technique as in fig. 1, A.)

c, male kidney (18 May), showing cholesterol-positive lipids in basal epithelium (arrowed) and aggregations of the same material in the tubule lumen. (Technique as in fig. 1, A.)

D, male kidney (18 May), showing depletion of lipid elements. (Technique as in fig. 1, A.)

E, male kidney (18 September), with lipid-free ‘sexual-segment’, but with apparent renal dements (arrowed) pronouncedly lipoidal and cholesterol-positive. (Technique as in fig. 1, A.)

F, female kidney (20 June), showing an appreciable steatogenesis in the larger elements which probably correspond with the male ‘sexual segment’ shown in fig. 2, c, D, E. (Technique as in fig. 1, A.)

31 March (6 specimens)

Seminiferous tubules

These were up to 260 μ in diameter, according to the individual. Spermatozoa were already present in great numbers and mitotic activity occurred throughout the germinal epithelium. Even basal spermatogonia were in division. In the more sexually advanced males spermatids occurred in the sub-peripheral row of cells. A dense concentration of lipid material had arisen near the tubule peripheries (fig. 1, A). A less profuse accumulation occurred in the central areas, which were often still without germ-cells. Such tubule lipid was generally weakly positive for cholesterol but was occasionally negative. In wax sections large vacuoles, from which’ lipids had been removed, were obvious in the spermatogonia. The basal lipids, therefore, did not come principally from the Sertoli cells nearby.

Leydig cells

These were arranged singly (as viewed in cross-section) or m tracts of up to 20. Individually they measured génerally some 12 μwide with a nucleus about 6 μ in diameter, but in some testes they were as big as 14 μ (nucleus 8 μ). The cytoplasm of many was densely sudanophil, but in some the remaining lipids were restricted to large, separate globules. In others, interstitial lipid had almost disappeared (fig. 1, A)—an index of the seasonal expenditure of the contained cholesterol. All interstitial lipid was heavily ‘iiolesterol-positive. In addition there had arisen numbers of small Leydig Ils which measured about 8 μ in diameter (nucleus 4 μ). These were non-lipoidal.

Epididymides

These contained massed spermatozoa in some animals but were still almost empty in others.

1. Less advanced group

Seminiferous tubules

These were up to 300 μ in diameter and contained great numbers of spermatozoa. Spermatogenesis was still proceeding at the tubule peripheries. At the same time a dense aggregation of lipid material was retained at the peripheries and some still occupied the central areas, which were now devoid of germ-cells. The peripheral lipids were sometimes Schultznegative; the central material was weakly positive.

Leydig cells

As in March, except that in general a greater seasonal depletion had occurred, even though the cells remained of the same size. A wide range of variation in lipid content occurred in the same organ. The remaining lipids were always heavily positive for cholesterol. Considerable numbers of small non-lipoidal cells also occurred.

Epididymides

All contained abundant spermatozoa. Some, however, appeared to have partly shed their contents.

2. More advanced group

Seminiferous tubules

Post-spermatogenetic collapse had begun and the diameter was reduced to about 160 μ. The centre of each tubule was occupied by a dense aggregation of lipid (fig. 1, B) that was weakly positive for cholesterol. An irregular peripheral zone of lipid material was still retained and it only sometimes gave a positive Schultz reaction. The epithelium was now confined to about five layers, only a few of which were in division. The current spermatogenetic rhythm had almost ended.

Leydig cells

Many groups of heavily lipoidal and cholesterol-positive interstitial cells were present (fig. 1, B). Some vacuolated, non-lipoidal cells still remained. Greater numbers of smaller cells had arisen but these as yet lacked cytoplasmic lipids.

Epididymides

These contained relatively few spermatozoa but, in addition, some lymphocytes. Those of some vipers were now almost empty again (e.g. the animal depicted in fig. 1, B).

18−22 May (9 specimens falling into three categories)
  1. Similar to less advanced group (category 1) of April except that two j limais had an apparently completely exhausted interstitium in which it was impossible to demonstrate cholesterol.

  2. Similar to the more advanced group (category 2) of April, except that, in addition, some spindle-shaped fibroblasts of the interstitium had become cholesterol-positive and so would apparently become part of the endocrine complement of the organ. Some exhausted Leydig cells now had pycnotic nuclei.

  3. Most advanced category, as follows:

Seminiferous tubules

These were reduced to a width of 160 to 180 μ according to individual animals. Few spermatozoa remained except for irregular clusters here and there. Aggregations of spermatids were still present but mitotic activity had stopped (fig. 1, c). In two animals not a single mitosis was observed among the spermatogonia. All tubules still retained varying quantities of lipid material in the peripheral regions and elsewhere.

Leydig cells

Clusters of mature, heavily lipoidal cells occurred. It was impossible to determine whether some of these constitute unexpended elements of the former sexual rhythm, or whether, more probably, all had quickly matured from the numerous juvenile Leydig cells, some of which still remained immature. However, no phases transitional between the two were observed. Also in the interstitium were groups of Leydig cells that still remained intact after exhaustion of their cytoplasmic lipids (fig. 1, D). In one testis a single mature Leydig cell was dividing (prophase), and in another animal several such phases were observed.

Epididymides

Some still contained spermatozoa; others were almost empty except for lymphocytes.

31 May (5 specimens)
Seminiferous tubules

Spermatozoa had disappeared. The tubules had increased slightly to about 200 μ in diameter and now contained dividing spermatogonia, primary spermatocytes, and secondary spermatocytes. The next season’s spermatogenetic rhythm had already begun. The tubule lumina were lined with lipid material and basal lipids also remained. The cholesterol reaction was faint and sometimes absent.

Leydig cells

In some vipers the initial stages of spermatogenesis had begun in the presence of only a very small number of mature Leydig cells. Juvenile Leydig cells were numerous, but were still only meagrely lipoidal. One mitosis was observed (fig. 1, E).

Epididymides were empty except for a few lymphocytes

30 fine (5 specimens)

Seminiferous tubules

Spermatogenesis was again proceeding strongly with iciany rows of germ-cells (including numerous secondary spermatocytes) almost filling tubules measuring about 200 μ, wide. Lipids, faintly cholesterolpositive, were now concentrated in the centre of each tubule except for a faint suffusion at its periphery. Thus sudanophil material was practically absent in the area of pronounced spermatogenetic activity.

Leydig cells

More and more were manufacturing lipids in the form of large, strongly cholesterol-positive globules, but some still remained non-sudanophil. Each sort measured about 10 μ in diameter (nucleus 5 μ). Smaller, non-lipoidal cells with a nucleus 3 μ in diameter were also common.

Epididymides were empty
24 July (9 specimens)

Seminiferous tubules were as in June, except that some animals already-contained spermatids and had lost all tubule lipids (and therefore cholesterol), apart from a few fragments in the narrow central lumina. Some vipers had whole tubules filled with actively dividing germ-cells (fig. 1, F).

Leydig cells

Here, too, remarkable transitional events were obvious: in some animals the interstitium was still only meagrely lipoidal (yet numerous germ-cells had arisen), while in others (in which the tubule lipids had largely vanished) the Leydig cells had become heavily lipoidal and aggregated into tracts as large as 15 by 10 μ (fig. 1, F). These were strongly cholesterol-positive.

Epididymides were empty.

12−18 September (8 specimens)

Seminiferous tubules were up to 240 μ in diameter according to the individual. Some already contained a few immature spermatozoa. In most animals tubule lipids had disappeared except for small quantities of cholesterol-negative fragments restricted to the centres. In some they were absent altogether. In two individuals, however, a minor autumnal steatogenesis had begun. The resultant fatty material, which was weakly positive for cholesterol, showed faintly in the peripheral regions and more strongly in the centres.

Leydig cells were at nearly maximum size (see under March). Many were losing their lipids and becoming vacuolated. All cytoplasmic lipid material was powerfully Schultz-positive.

Epididymides were empty

28 October (2 specimens)

Seminiferous tubules were up to 220 μ wide and packed with germ-cells, including a few early stages of spermatozoa. Again faint traces of an autumnal steatogenesis were discernible.

Leydig cells were heavily lipoidal, and cholesterol-positive aggregations occurred. The cytoplasm of these, however, showed no sign of seasonal depletion.

Epididymides were empty.

18 November (5 specimens)

Seminiferous tubules were as in October, but with slightly increased steatogenesis, a greater number of spermatids, and a few fully formed spermatozoa.

Leydig cells were as in October except that several mitotic divisions (prophase) were seen in one animal.

Epididymides

Despite the presence of tubule spermatozoa the epididymides were empty.

3 January (4 specimens taken from experimental hibernation)

Seminiferous tubules were up to 230 μ in diameter, but it cannot be suggested that the minor apparent increase had occurred during hibernation, because spermatogenesis in each animal was at a halt. The most advanced stage was that of spermatid (fig. 2, A). The tubules were faintly suffused with lipid that was very dubiously positive for cholesterol.

Leydig cells

These were heavily charged with lipid, including cholesterol, and were 12 μ in diameter (nucleus 6 μ).

Epididymides were empty

Juveniles taken on 30 June (3 specimens)

Seminiferous tubules were between 50 and 90 μ wide, with spermatogonia in division and primary spermatocytes.

Leydig cells were in tracts up to 70 by 50 μ, the individual cells being 10 μ wide (nucleus 5 μ). They were lipoidal and cholesterol-positive (fig. 2, B).

Epididymides were empty

Like those of birds, the testes of the Chelonia, Crocodilia, and Lacertilia contain Leydig cells of seasonally-variable lipid content and tubules that undergo a post-spermatogenetic steatogenesis. Therefore, however far modern snakes (e.g. Vipera) have changed from the ancestral archeosaurian stock from which sprang the Aves, it is probable that such characteristics are derived in both classes from a common ancestry. As might be expected, a comparison, between the cyclical events in the testis of birds and viper shows both similarities and differences.

The points of similarity are :

1. As distinct from the minor lipophanerosis that occurs typically at the end of vertebrate spermatogenesis, the unshed germinal material of both reptiles and birds metamorphoses into a mass of cholesterol-positive lipids. This material gradually disappears, but not down the vas. The cholesterol may be of endocrine significance ; see, however, below, in regard to Vipera.

2. With the heightening of spermatogenesis and sexual behaviour, the cholesterol-positive lipids of the interstitial Leydig cells are gradually depleted, presumably contributing to the manufacture of the male sex hormone. In both reptiles and birds a new, lipoidal interstitial generation arises after the seasonal exhaustion of the old.

3. Autumnal sexuality occurs in some vipers as also in many avian species, as a consequence of the post-nuptial development and maturation of the interstitium.

Related to the above described common pattern are the following differences in detail:

1. The tubule lipids of the viper are only weakly positive for cholesterol and soon after their genesis become almost or wholly negatively so, whereas neighbouring Leydig lipids, whenever present, are always powerfully positive. In birds both tubule and interstitial elements are equally and strongly cholesterol-positive. It is impossible in the viper, therefore, to suggest with any assurance that such briefly-appearing, meagre cholesterol is as functionally important as it may be in birds.

2. In the viper spermatogenesis starts and continues while the tubules are still clotted with post-nuptial lipids, so that for only a brief transitional period during autumn is the testis almost without tubule-lipids. In birds such lipids remain for a lengthy period, varying between species, but always disappear abruptly at the onset of the next spermatogenesis, so no trace of them remains by the time large numbers of primary spermatocytes appear. In the viper (which begins its next spermatogenetic rhythm almost immediately after the preceding one ends), spermatogenesis proceeds strongly beneath a heavy cargo of tubule fat, which has, however, almost disappeared by the time spermatids have appeared.

3. In the viper the early stages of spermatogenesis (at least up to primary spermatocytes in synizesis) are sometimes achieved while the newly regenerated Leydig cells are still only meagrely lipoidal and at the most only weakly cholesterol-positive. The new interstitium sometimes does not become heavily lipoidal, cholesterol-positive, and mature until several rows of new germ-cells have arisen (in mid-summer). In birds, on the contrary, the new spermatogenesis occurs only after the new interstitium has become relatively mature.

4. In Vipera there exist relatively few Leydig cells, but two means (each extremely rare among birds) of increasing their numbers exist. In birds mitotic division of mature Leydig cells is rare: in Vipera it is not uncommon, having been occasionally observed in March (Volsoe), May (fig. i, E), and November. A second means of increasing the endocrine complement of the testis of the viper is the assumption of a lipid and cholesterol-positive cytoplasm by odd spindle-shaped interstitial fibroblasts during the post-nuptial period of rehabilitation.

5. Almost immediately after the spring shedding of spermatozoa there begins in Vipera the succeeding spermatogenetic rhythm—a phenomenon unknown in birds. A broadly comparable rapid recovery of spermatogenetic function occurs, however, in the pike, Esox lucius (Lofts and Marshall, 1956) and in the common frog, Rana temporaria (van Oordt, 1956). This is almost certainly an adaptation related to the poikilothermie condition and to the thermal fluctuations of the temperate environment. The rapid renewal of spermatogenetic processes ensures that in the viper many secondary spermatocytes, a few spermatids, and often some spermatozoa have been formed before the late autumn hibernation. This is a normal and unvarying component of the sexual cycle, whereas when autumnal spermatogenesis occurs in birds it is wasteful: it generally leads only to a second lipoidal metamorphosis or, occasionally, to an almost invariably unsuccessful winter reproduction. But in Vipera the summer spermato6enesis stops during hibernation and, when the snake emerges in the spring, the sexual cycle, already at an advanced stage, begins anew. At the same time fertilization in the viper does not occur earlier than that of most birds, most of which contain merely spermatogonia at the end of January. The high rate of avian metabolism enables birds to achieve in the space of a very few weeks events that in Vipera occupy months. The rook (Corvus frugilegus), for example, does not begin its annual spermatogenesis until very late in January or early in February, but it copulates by mid-March (Marshall and Coombs, 1956). Vipera, on the contrary, emerges when temperatures rise to about 8° C (Volsoe, 1944), basks in the sun as it completes its spermatogenesis, and copulates in April and May.

6. In the viper there is only a minor variation in tubule-diameter and, it follows, total testis size. Similarly, there is apparently no marked and sudden seasonal renewal of the tunica albuginea such as occurs in birds and in the pike. At the same time, however, we saw a tunic variation from between 5 and 20 μ in width, and, in some specimens taken from April to July, noted a tendency for the coat to be separated into two layers. The inner and thinner (from 5 to 10 μ, thick) appeared to be more densely composed and so it is possible that in this animal too, some tunic rehabilitation occurs at the time when other regenerative events are taking place.

The present investigation, although primarily concerned with the seasonal distribution of testis lipids, has served incidentally to confirm many aspects ol the work of Volsoe (1944). We and he are at variance on only two points. I hus, Volsoe writes of the scarcity of interstitial cells in Danish vipers, but we have not found them rare, although it is true that they are fewer than in any bird so far investigated. The supposed rarity and low degree of variation in interstitial cells is seen to be an illusion if we correctly identify the rounded, initially non-lipoidal juvénile Leydig cells (apparently of connective tissue origin) and, by means of Sudan reagents, follow the seasonal depletion oi cholesterol in their cytoplasm after maturity. These, sometimes at least, become pycnotic. The apparent lack of variation is engendered by their capacity for very suddenly reaching their almost, or quite, maximum size and, as far as we have seen, the persistence of at least some mature cells in the interstitium during hibernation and perhaps (unlike in birds) even after the conclusion of spermatogenesis.

Secondly, we have been unable to confirm with British material Volsoe’s statement that ‘apart from the pause which occurs during hibernation, spermatogenesis is continuous, one spermatogenetic cycle beginning even before the preceding is concluded’. Although, in a given British population, we found individuals still producing the last spermatozoa of the season, and other animals already with the first mitotic spermatogonia of the next rhythm, we found beside them still other vipers with reduced tubules containing a few remaining spermatozoa and aggregations of inactive spermatids—yet with no mitotic divisions whatever (fig. i, c). The basal layers of such animals were composed of rows of spermatogonia which showed no sign of modification. Therefore, we believe that in some British vipers at least there is a distinct, though brief, lull between the end of one spermatogenesis and the start of the next. Very quickly, however, the next cycle begins: the poikilothermous viper almost immediately takes advantage of the spring and summer warmth. It completes much of its next spermatogenesis before the females have dropped the young of the preceding rhythm.

Although autumn mating in vipers has been observed in Guernsey, and combat in October in Surrey (Smith, 1951), Volsoe found no pregnancy among naturally hibernating females in Denmark. Of the 5 adult females put by us into hibernation on 30 September, one had produced young and a second contained embryos of about birth-size when the torpid snakes were killed on 30 January. These results, however, cannot be said to support the belief of Wolleback (cited and rejected by Volsoe), based on the discovery of a pregnant female in Norway late in March, that young may be born during natural hibernation. From the size of our January embryos and young it is perhaps possible that autumn sexual behaviour might have led to pregnancy just before their enforced hibernation. It seems far more probable, however, that our interference retarded the normal autumn parturition and that the females were still gravid from a spring mating when put into darkness. Both Vainio (1931) in Finland and Bernstrom (cited by Volsoe, 1944) in Sweden showed that in higher latitudes there exists an adaptation that allows gestation to continue through the brief northern summer into the next hibernation, so that reproduction occurs every second year.

As with birds of the temperate zone, the sexual processes of Vipera are inhibited by winter conditions, but in the viper the winter cessation of spermatogenesis is not followed by profound retrogressive changes. The few autumn spermatozoa may degenerate, but the relatively advanced germinal stages retain their integrity and the cycle appears to begin without further delay upon emergence from hibernation. The epididymides of some Scottish vipers were full of massed spermatozoa on 31 March, somewhat earlier than in Denmark. On the other hand, neither Volsoe nor ourselves found evidence of free spermatozoa during the autumn. He found them in the epididymides until July. In our snakes they disappeared by the end of May.

Volsoe’s information that vipers emerge from hibernation when air temperatures reach about 8° C, and a consideration of their subsequent predilection for basking in the sun make it difficult to resist the conclusion that the sexual cycle is essentially under thermal control. It is obvious, however, that other factors are involved. Spermatozoa in our animals had been shed (i.e. the epididymides were empty) by the end of May and some spermatids of the next generation were formed by 24 July. In many vertebrates, including apparently the viper in spring, the concluding stages of the spermatogenetic cycle occupy only a matter of a few weeks or, in birds, even days. Although the days remain hot and relatively long between July and September, very few spermatozoa are formed by the time the vipers, now again probably controlled by air temperature, return to hibernation in November. Clearly there has been some intervening resistance to spermatogenesis. As a consequence, and despite widespread autumnal sexual behaviour, fertilization (whatever the condition of the female) cannot often take place at the beginning of the winter.

Because of the reported reproductive significance of the ‘sexual segment’ of the renal tubule, we tested sections of the kidney for lipid and cholesterol. We found the epithelium of the wider tubules positive to the Schultz test, but were unable to correlate the observed cyclical changes with events in the interstitium. We found, in fact, that reptiles taken at the same period and place varied considerably in regard to their renal condition, even when the testis elements were in an approximately identical state. It is important to emphasize that much of our material was not gathered by ourselves, and that (despite guarantees to the contrary by a professional collector) some of it may occasionally have remained some time in captivity before dispatch to the laboratory. We found, however, that in the spring the epithelium and lumen ol the ‘sexual segment’ could be heavily spotted with lipids and cholesterol (fig. 2, c) when the Leydig cells were either essentially non-lipoidal (fig. 1, A) or heavily so (fig. 1, B). Similarly, there was no clear-cut correlation between the condition of the sexual segment and that of the seminiferous tubules, although there was a suggestion in some animals that an expenditure of sexual segment cholesterol occurred during both the spring and late summer periods °f spermatogenetic activity.

It was at least certain, however, that a cycle of activity takes place in the renal epithelium itself. Fig. 2, c shows a large ‘sexual segment’ (arrowed), mottled with lipid droplets (containing cholesterol); and fig. 2, D depicts that °f a viper, received and killed on the same day (18 May), vacuolated and depleted of such substances. Fig. 2, E shows another phenomenon not previously commented on: after the clearing of the ‘sexual segment’, there has occurred a marked steatogenesis (arrowed) of the smaller and apparently essentially ‘renal segment’ of the tubule.

The lipids thus indicated in fig. 2, E were as heavily laden with cholesterol as were those of the ‘sexual segments’ of fig. 2, c. In the female, too, relatively minor changes, involving a similar steatogenesis (fig. 2, F) take place in that part of the tubule which probably corresponds with the sexual segment of the male. The problem of the cyclical changes in the kidney in relation to external factors and the gonads invites an experimental analysis with animals from a rigidly controlled environment.

We are grateful to Dr. Avrion Mitchison for help in obtaining material. One of us (F. M. W.) was in receipt of a grant from the Central Research Committee of the University of London.

Coombs
,
C. J. F.
, and
Marshall
,
A. J.
,
1956
.
J. Endocrin
.,
13
,
107
.
Gampert
,
O.
,
1866
.
Zeit. wiss. Zool
.,
16
,
369
.
Lofts
,
B.
, and
Marshall
,
A. J.
,
1956
.
Quart. J. micr. Sci
.,
98
,
79
.
Marshall
,
A. J.
,
1949
.
Quart. J. micr. Sci
.,
90
,
265
.
Proc. Zool. Soc.
Lond
.,
121
,
727
.
Mem. Soc.
Endocrinol
.,
4
,
75
.
and Coombs
,
C.
J. F
.,
1956
.
Personal communication
.
and
Lofts
, B
.,
1956
.
Nature
,
177
,
704
.
Morley
,
A.
,
1943
.
Ibis, §5, 132
.
Oordt
,
P. G. W. J. van
,
1956
.
Regulation of the spermatogenetic cycle in the Common Frog (Rana temporaria). Arnhem (Vander Weil)
.
Regaud
,
C.
, and
Policard
,
A.
,
1903
.
Arch. Anat. micr
.,
6
,
191
.
Serventy
,
D. L.
, and
Marshall
,
A. J.
,
1956
.
Emu, in press
.
Smith
,
M.
,
1951
.
The British amphibians and reptiles. London (Collins)
.
Vainio
,
I.
,
1931
.
Ann. Soc. Zool. Bot. Fenn
.,
12
,
1
.
Volsoe
,
H.
,
1944
.
Structtire and seasonal variation of the male reproductive organs of Vipera berus (L.). Copenhagen (Bianco Lunos Bogtrykkeri A/S)
.