^*Studies on the Germ-cell Cycle of Cryptocotyle lingua: II. Germinal Development in the Larval Stages

Investigations of the germ-cell cycle in the digenetic Trematoda, especially the more recent studies of Brooks (1930) and Woodhead (1931), have resulted in interpretations so opposite in nature that further study of this long-disputed question has become increasingly desirable. Previous explanations have, for the most part, been based on studies dealing with only the larval stages. For that reason, an interpretation of the entire lifecycle with reference to the concept of germinal continuity, as applied to trematodes, has been virtually impossible.

A comprehensive study of the trematode germinal cycle involves some of the most difficult as well as fundamental problems of parasitology. Although many hypotheses have been advanced, it is not yet clear how the digenetic trematodes have acquired the present complicated life-histories. Was the vertebrate or the molluscan host the first to be parasitized ? If the mollusc, did the early trematode become sexually mature in that host, thereby completing the life-cycle in a manner similar to the freeliving and ectoparasitic flatworms, or did it immediately acquire one or more intercalary stages similar to those in the present molluscan host ? If the first host was a vertebrate, how did the mollusc and other intermediate hosts become involved? To what extent has germinal development changed with the evolution of the parasite and modification of the life-cycle ? Are the interrelationships of various trematode families sufficiently close to expect the group as a whole to be characterized by the same type of germ-cell cycle ? Are there to be found among free-living or parasitic turbellarians, or in the life-histories of the Monogenea, any stages or modes of development such as polyembryony, internal budding or parthenogenesis, that foreshadow or correspond to the intercalary stages of the Digenea ? If not, are we justified in assuming then that endoparasitism alone is responsible for these larval stages ?

It is believed that satisfactory answers to these questions must come from careful and precise studies dealing not only with the entire cycle of a number of representative digenetic trematodes, but also with other forms which are evidently related to this group, particularly the rhabdocoele turbellarians, monogenetic trematodes, and the Mesozoa (Rhombozoa).

The morphology and parasitic habit of certain of the rhabdocoele turbellarians are suggestive of the Trematoda. In form and habit, Anoplodium, an endoparasite of holothurians, and Graffilla, from the mantle chamber of molluscs, have been regarded as similar to the ancestors of the Digenea and Monogenea respectively. The life-history of Anoplodium, however, is direct, and larval stages similar to those of the Digenea are lacking.

In development and life-history, similar phenomena have been described for both turbellarians and trematodes. Polyembryony, which has been reported for both monogenetic and digenetic trematodes, has also been described for certain of the turbellarians in which two or more embryos (two or three in Graffilla to as many as thirteen in Syndesmis) develop in a single egg capsule. Linton (1910), in his studies on Graffilla gemellipara from the mussel, Modiolus plicatulus, believed that the two or three embryos contained in each capsule came from a single fertilized ovum. This turbellarian was probably mistaken by Nicoll (1906) for a trematode sporocyst, as pointed out by Patterson (1912) who concluded, in regard to its development, that the twin embryos result from the enclosure of two fertilized ova in each capsule and not from the division of a single ovum or embryo.

There are features in the life-cycle and germinal development of Dicyema (Whitman, 1882; Hartmann, 1925), and perhaps other Mesozoa, which may throw considerable light on the interpretation of the digenetic cycle. The Mesozoa are chiefly parasites of cephalopods, although orthonectids infest turbellarians, nemertines, annelids, and echinoderms. The group as a whole has been regarded by certain zoologists as intermediate in position between the Protozoa and the Metazoa. The bulk of evidence indicates, however, that the Mesozoa are flatworms which have become degenerate as a result of parasitism. Certain stages of their life-cycles resemble trematode miracidia and sporocysts, particularly with regard to their germinal development, as pointed out by Reuss (1903). The early cleavage stages of the mature agametes of Dicyema, and of the mature germ-cells of Cryptocotyle rediae, are quite similar in nature. These and other cytological resemblances will be discussed later in connexion with germinal development of the trematodes.

Although the relation of the monogenetic trematodes to the Digenea is a matter of considerable controversy, and is beyond the scope of the present investigation, the life-histories and modes of reproduction in certain monogenetic species are of interest to this study. The developing young of Gyrodactylus elegans were observed by Wagener (1860) and other early investigators to contain embryos of a third generation while still within the body of the parent. Von Linstow (1892) believed that the secondary embryos resulted from asexual reproduction in the primary or daughter embryo. Metschnikoff (1870), however, insisted that the secondary embryo has the same origin as the primary, namely, the segmentation products of the fertilized egg ; the primary and secondary embryos would accordingly be interpreted as ‘sisters’ rather than ‘mother and daughter’, since they are of the same age. Kathariner (1904) supported Metschnikoff’s opinion and, in a later paper (1920), applied his findings on the monogenetic species, Gyrodactylus elegans, in a review of life-histories and germinal continuity in the digenetic trematodes.

As a rule, endoparasitic trematodes have digenetic life-cycles while ectoparasitic forms are monogenetic. Certain of the Aspidocotylea, however, are exceptions to this generalization in that they are true endoparasites and, as far as is known, are monogenetic in development. Species of Aspidogaster have been described from the coelomic cavities of molluscs and the intestine of fishes and turtles. It seems quite possible in this case that Aspidogaster is typically parasitic in bivalves, and is merely carried over into the vertebrate when the mollusc is eaten. It is further possible that the occurrence of Aspidogaster in vertebrates represents a case of incipient endoparasitism. Leuckart, however, regarded the aspidogastrids as sexually mature rediae whose former adult stages had been eliminated.

The relation between habitat and life-cycle is further complicated by the observations of Gallien (1932) on Polystomum integerrimum, whose developmental cycle, according to him, consists of two distinct generations. The first is ectoparasitic on the gills of young tadpoles and produces a second generation which becomes attached to the gills of older tadpoles. Upon metamorphosis of the host, the polystomes pass down the alimentary canal and become established as endoparasites in the urinary bladder of the frog. This is a case, therefore, in which an apparently digenetic life-cycle, consisting of an ectoparasitic and an endoparasitic generation, involves the larval and adult stages of a single host species.

While some of the earlier investigators regarded the Monogenea as close to if not directly in the phylogenetic line of the Digenea, Leuckart’s opinion concerning the Aspidocotylea would derive that group at least from the digenetic trematodes. Since the Monogenea are largely ectoparasites of vertebrates and have direct development, it is possible that they adopted the parasitic habit at a comparatively recent time, after the appearance of vertebrates, and therefore that they are not closely related to the older digenetic group.

The present paper completes, as far as is possible with the material at hand, a study of the germ-cell cycle in Cry pt ocotyle lingua Creplin, a digenetic trematode belonging to the family Heterophiidae ; gametogenesis in the adult has been described in an earlier report (Cable, 1931). Specific investigations to determine the origin and differentiation of germ-cells, the occurrence of maturation or other sexual phenomena, and germinal continuity in the larval stages have been undertaken. Stunkard (1930), in his studies on the life-history of Cryptocot y 1 e, did not succeed in hatching the eggs, and stated the belief that unhatched eggs are eaten by the snail. Since the writer has not had an opportunity to test this hypothesis, one stage of the life-history, the miracidium-mother-sporocyst, has not been observed. It is surprising that at least recognizable fragments of this stage were not found in the dozens of infected snails that were carefully examined in serial sections during the course of the study.

The life-history of Cryptocotyle may be described briefly as follows. Thè adults inhabit the small intestine of fish-eating birds and mammals. Feeding experiments have demonstrated a low grade of specificity for these hosts. Eggs pass out with the faeces, and if they reach sea-water, miracidia develop within about ten days. The marine snails, Littorina littorea and Littorina rudis, serve as molluscan hosts, in which there is a single redial generation. Cercariae escape from the snail and encyst in the skin of the cunner, Tautogolabrus adspersus, and other fishes. When the.infected fish is eaten by a suitable bird or mammal, the cyst-wall is ruptured and the young -worm develops to sexual maturity in about five days.

The writer expresses his grateful appreciation of the ever-willing encouragement and advice of Doctor H. W. Stunkard, who suggested and directed this study.

Many theories and interpretations of the germinal cycle of digenetic trematodes have been advanced, but none of them has become generally accepted. Brooks (1930) classified these hypotheses as follows: metagenesis, heterogeny, paedogenesis, extended metamorphosis and germinal lineage with polyembryony. To this list may be added Woodhead’s (1931) concept of polymorphism with three sexually mature hermaphroditic generations.

The theory of metagenesis, or alternation of a sexual with an asexual stage, was applied to the trematodes by Steenstrup (1842), who likened their development to that of the coelenterates and tunicates. This view was supported by Moulinié (1856), Pagenstecher (1857), Wagener (1866), Balfour (1880), and Biehringer (1884). Since the first trematode life-history, that of Fasciola hepática, was not traced experimentally until 1883, the opinions of these earlier investigators were based on homologies of larval and adult stages.

Grobben (1882) first suggested that heterogeny (alternation of a bisexual with a parthenogenetic generation) occurred in the Digenea, and expressed the belief that cercariae develop from parthenogenetic ova. The germ-cells of sporocysts and rediae have been the object of several more recent investigations, the point in question being chiefly to determine whether or not polar body formation or other maturation phenomena are associated with these cells. Reuss (1903) observed, in the sporocysts of Distomum duplicatum, that the ‘ovum’ was accompanied by three smaller cells which he interpreted as polar bodies. Haswell (1903), in his studies on two sporocysts occurring in Mytilus latus, stated in a footnote that he did not confirm the findings of Reuss, but observed that ‘in a large proportion of the specimens there occurs lying loose in the sporocyst in the immediate neighbourhood of the ovary a varying number of cells (fig. 14) which have homogeneously deeply staining nuclei 0·002 mm. in diameter. If these are not of the nature of polar bodies it seems difficult to account for them’. Haswell also reported that the ovary, after considerable division of its component cells, became fixed in position and completely enclosed in a delicate membrane. The ova gave rise to both cercariae and daughter sporocysts. Upon leaving the mother sporocyst, the daughter sporocysts, often containing cercarial embryos, were observed to divide by binary fission.

Tennent (1906) reported finding polar bodies in the sporocysts of Bucephalus haimeanus, from the gonads of the oyster. He described two methods of germ-cell origin, both of which were from the wall of the sporocyst. In the first, there seemed to be no definite place of origin ; in the second, germ-cells were produced by a definite ‘keimlage’. In the older germ-tubes of the branching sporocysts, cell-boundaries disappeared and the cytoplasm condensed around the nuclei, forming germ-cells which passed into the lumen of the tube. In a few cases, Tennent described the cutting off of a small cell, after which the nucleus moved to the periphery of the larger cell and divided unequally to form a second small cell. Cleavage followed this process, and the germ-balls were of considerable size when they left the keimlage and entered the lumen of the sporocyst. Tennent also reported finding central opaque masses in the young sporocysts.

Cary (1909) studied germinal development in a sporocyst which he thought was that of Diplodiscus temporatus. It has since been shown by Cort (1915) that Cary confused two distinct species, but it is of interest to note that in the form with which he worked Cary reported that germ-cells developed from cells of the body-wall. In the process of maturation, a single polar body was extruded without a reduction of chromosome number. Faust (1917) reported the occurrence of true parthenogenesis in the rediae of Cercaria flabelliformis. He observed that the ‘ovum’ extruded a single polar body which sometimes divided. Chromosome reduction did not occur. He refuted the arguments of Rossbach against polar body formation, stating, “The polar bodies have been found not only in cytoplasmic continuity with the ovum, but in the actual stage of mitosis preceding the separation of the polar nucleus from the germ-ball”. He described a specialized germinal mass at the blind end of the gut as the source of germ-cells. In simpler types such as Cercaria diaphana and Cercaria micropharynx, he reported that the entire layer beneath the epithelium was germinal in nature.

A large number of workers, on the other hand, have been unable to find evidence of maturation in larval trematodes, and hence maintain that heterogeny is not characteristic of this group. Looss (1892) did not report the occurrence of polar bodies in the larval stages of Amphistomum subclavatum. In his figures are shown small, densely-staining nuclei in the germballs which he designated as “degenerierende Kerne (?) der Keimballen”. The degeneration of entire germ-balls is also figured by Looss. He described the production of germ-cells both from a definite keimlage and from cells of the body-wall. In the latter method the cells of the wall divided, one of the daughter cells remaining in place, the other being cut off into the lumen where it developed into a germ-ball.

Coe (1896) found no polar bodies in the sporocysts and rediae of Fasciola hepática. His material, however, was unfavourable for study and he preferred to draw no conclusions. Eossbach (1906), working on Cercaria armata and Cercaria echinata, observed cells similar to those interpreted as polar bodies by Eeuss (1903). Rossbach, however, concluded that these small cells belonged to the soma of the sporocyst or redia, and had nothing to do with the germinal development. They were not in direct continuity with the germ-cells, present often in larger numbers than threes, and were found in all stages of development, including the ovary of sexually mature adults. Mathias (1925), in his studies on the developmental cycles of Strigea tarda, Hypoderaeum conoideum, and Psilotrema spiculigerum, did not observe polar bodies or other maturation phenomena. Dollfus, in a private communication to Brooks (1930), stated that he had never found indisputable polar bodies in the thousands of sporocysts and rediae which he had examined. Among the material investigated by Dollfus were the larval stages of Bucephalus haimeanus, the same species for which Tennent (1906) described maturation phenomena. Dubois (1929) disagreed with Dollfus, claiming to have observed germinal discontinuity and parthenogenesis in the sporocysts of Cercaria helvética V.

The concept of a metamorphosis extending over several generations was supported by Balfour (1880), Leuckart (1886), and Looss (1892). Balfour outlined the theoretical life-history of a typical trematode, stating that the majority of its stages “are simply parts of a complicated metamorphosis, but in the coexistence of larval budding (giving rise to Cercariae or fresh Rediae) with true asexual reproduction there is in addition a true alternation of generations”. Looss (1892), in expressing a similar opinion, homologized the sporocyst, redia and cercaria, pointing out that they possess the same general structure. The concept of extended metamorphosis was strengthened by the early studies on the monostomes in which the miracidium contains a single fully formed redia at the time of hatching, and, for that reason, was likened to Desor’s larva of the nemertines. This condition has also been described by Linton (1914) for Parorchis avitus, a distome species from gulls.

Germinal lineage, both with and without polyembryony, has been applied to the trematode life-cycle by a number of workers. Leuckart (1886) believed that the germ-cells of sporocysts and rediae are derived directly from the egg, and remain in an undifferentiated condition in the body-cavity of sporocysts and rediae during their somatic development. Various modifications of this theory have been accepted by Thomas (1883), Schwarze (1885), Coe (1896), Dollfus (1919), Kathariner (1920), and Brooks (1928, 1930). From his studies on the larval stages of several distome and monostome species, Dollfus (1919) concluded that sporocysts, rediae, and cercariae do not develop from the wall of sporocysts and rediae, but from one and the same germinal line which comes from the segmentation of the fertilized ovum. The cells of this line give rise to the larval somatic tissues of the rediae and sporocysts by a sort of internal polyembryony. These somatic tissues are merely larval envelopes which have enclosed the cells of the germinal line and, being sterile, take no part in reproduction. Kathariner (1920) also supported the theory of germinal lineage, but maintained that polyembryony in sporocysts and rediae was only apparent since the germ-cells were not derived from but merely enclosed by the soma of these stages. The germinal line would accordingly be uninterrupted from fertilized egg to fertilized egg.

From his studies on the larval development of twenty species of trematodes, Brooks (1930) interpreted the life-cycle ‘as being one where the germinal lineage passes through successive larval stages in which polyembryony features as a mode of multiplication and in which precocious cleavage of the germ-cells is the activating factor’. According to Brooks, the primordial or ‘antecedent’ germ-cells which are found in the body-cavity of the miracidium-mother-sporocyst divide to form loosely organized ‘germ-masses’. These dissociate and their components may by division form ‘secondary germ-masses’ or may develop into germ-balls. ‘Tertiary germ-masses’ may possibly be produced by the components of the ‘secondary germ-masses’. The dissociation of these masses was interpreted by Brooks as typical polyembryony. He maintained that sporocysts and rediae are homologous, the extent to which their somatic structure is developed depending upon the intensity of the factor for precocious cleavage. In the germ-masses, this factor is so strong that no observable somatic structure is formed, while in the development of daughter-sporocysts, rediae, and daughter-rediae, the tendency toward precocious cleavage being progressively decreased, somatic expression becomes correspondingly greater. The precocious tendency of the germcells eventually becomes decreased to such an extent that the embryos are able to undergo ‘natural’ development, producing larvae (cercariae) in which the germinal elements are normally restrained. Since polyembryony has been experimentally produced in a number of forms by lowered oxygen tension and changes in other factors such as temperature (Stockard, 1921) and food (Marchai, 1904), Brooks suggests that the cause of polyembryony in trematodes may likewise be exogenous in nature. Since a reduced oxygen supply may produce polyembryony, he reasons that this phenomenon may have occurred in the digenetic trematodes on adoption of the parasitic habit in which less oxygen was available than in the free-living state. Brooks did not find polar bodies or structures that could be interpreted as ovaries.

Polymorphism has recently been advanced by Woodhead (1931) as an explanation of the life-cycle in the gasterostomes. He interpreted the sporocyst and redia, which he has described for the first time in this group, as adults rather than larvae, and has described gametogenesis and fertilization in these stages as well as in the final adult gasterostome. Woodhead concluded, accordingly, that Bucephalus is polymorphic, the life-cycle consisting of three sexually mature adult generations. The sporocyst is regarded as a¹ dendritic’ colony. The redia develops as a ciliated ‘pro-redia’ which metamorphoses into a mature hermaphroditic redia. If the observations of Woodhead are correct, sexual phenomena may be expected to occur also in the larval stages of other trematodes supposedly not too distantly related to the Bucephalidae. La Rue (1926), in his revision of the taxonomy of trematodes having furcocercous cercariae, related the Bucephalidae to the strigeids and schistosomes. Studies on the germinal development of the latter two groups have not thus far tended to support Woodhead’s observations concerning the Bucephalidae. In a preliminary note, Woodhead (1932) reported that Leucochloridium, which is not generally regarded as being closely related to the Bucephalidae, also has a polymorphic life-cycle similar to that of Bucephalus. This observation adds considerable support to his earlier conclusions which, if correct, necessitate an entirely different approach from that hitherto taken in interpreting the trematode germinal cycle.

Dollfus (1929, 1932) has recently described cases of sexually mature metacercariae and has postulated that they may be progenetic, that is, retaining the degree of somatic differentiation characteristic of the larva at the time sexual maturity is attained. Dollfus (1932) suggested the possibility that both an abbreviated and a ‘normal’ life-cycle occur in one of the Lepodermatoidea having sexually mature metacercariae, although a vertebrate host has not been found.

With such a diversity of observations and opinions, it is apparent that the interpretations of trematode germinal development thus far advanced are far from satisfactory when applied to the group as a whole.

The larval stages of Cryptocotyle lingua in the liver of the periwinkle, Littorina lit torea, were used in the present study. The infected snails were collected by the writer during the summers of 1931 and 1932 at Woods Hole, Massachusetts. They were crushed and the livers having lighter infections were fixed in a number of fluids, including Flemming’s strong solution, with and without acetic acid, picro-formol, picro-formol-acetic, Allen’s B₃, with and without chromic acid, corrosive sublimate-acetic, and saturated aqueous corrosive sublimate. Flemming’s solution with acetic acid and picro-formol-acetic gave the best results, penetration being poor with most of the others employed. The material was imbedded in paraffin and sectioned at six and seven microns, such sections requiring less reconstruction than thinner ones and, at the same time, giving good cytological differentiation. Heidenhain’s iron alum-hematoxylin and safranin, both with and without counter stains, were employed, the former with eosin generally giving the better results. Living material was studied by teasing the larvae from the host tissue, compressing slightly with a no. 1 cover-glass on a slide, sealing with vaseline or paraffin to prevent evaporation, and examining with the oil-immersion objective. This method, however, was of little value in the study of germinal development. Fig. 11, Pl. 32, was made with the aid of a camera lucida; all other drawings were made from measurements free-hand and to the same scale. The photomicrographs of Plates 33-5 were made with a Zeiss camera, using Eastman (Wratten & Wainwright) panchromatic plates.

The miracidium-mother-sporocyst was not observed during this study although thousands of sections were carefully examined. Very young rediae, however, were found in considerable numbers in material having lighter infections. They were also present in livers having heavier and older infections, usually being located in the lymph spaces just beneath the epithelium of the large bile duct and in interlobular spaces of hitherto uninvaded regions of the liver. Their occurrence along with mature cercariae suggests that either multiple infection takes place or the mother-sporocyst persists until a late stage, producing rediae more or less continuously. If the latter is true, the mother-sporocyst is either so lacking in organization that it cannot be recognized as such, or remains in some part of the host other than the digestive gland where it produces large numbers of rediae which make their way into the liver to complete their development. Since it was expected that at least some of the livers collected would contain sporocyst material, the remainder of the snail was not prepared for study. The youngest rediae observed (fig. 12, Pl. 33) are about 100-25 microns in length and exhibit a low grade of differentiation, the muscular oral sucker and the gut being the only well-defined structures observable in sectioned material. Due to the abundance of connective tissue throughout the posterior half of the larva, the body-cavity is not well defined, but is indistinctly bounded by somewhat heavier strands of tissue which are roughly parallel to the bodywall. Cell boundaries are not visible and the nuclei of the body-wall, supporting tissue of the sucker, and the primordial germ-cells are indistinguishable with respect to size and staining reaction. This is clearly seen in fig. 12, Pl. 33. The germ-cells lying among strands of connective tissue within the region of the future body-cavity are distinguishable from all other cells of the body only by their position. At this stage, the large vesicular and the small densely-staining nuclei, characteristic of the supporting tissue of the sucker in older rediae, are all of the small darkly-staining type. The dense appearance of all the nuclei of the very young redia is only slightly affected by destaining which completely bleaches the surrounding host tissues.

The primordial germ-cells are believed to be distinct from somatic tissues when the young redia leaves the mother-sporocyst, since further cell divisions are not observed until germinal development is initiated. Mitotic figures are of extremely rare occurrence in the somatic cells of the redia at any stage. The nuclei of the primordial germ-cells (fig. 1, Pl. 32) are round or oval in outline and about 3·5 microns in diameter. The centre and periphery of the darkly-staining nucleus are intensely blueblack, indicating a higher chromatin concentration at these parts of the cell. Cytoplasm and cell boundaries are not distinct ; the nuclei appear to lie in light areas bounded by strands of connective tissue. Germinal differentiation exhibits a distinct anterior-posterior gradient, the more anterior of the germ-cells initiating the process. This development is characterized by increase in size, progressive nuclear changes, and condensation of cytoplasm with the gradual appearance of more distinct cell boundaries. As the germinal nucleus grows (figs. 2, 3, Pl. 32), the chromatin becomes more diffuse, appearing first as dark masses (fig. 2, Pl. 32) more or less evenly distributed throughout the lighter nucleoplasm. The karyosome appears as an irregular body near the centre of the nucleus. Further differentiation is characterized by continued growth and less intense staining. The chromatin appears as small knots on a reticulum which is denser near the periphery of the nucleus. The karyosome also increases in size, approaching 2 microns in diameter, becomes more spherical in form and is characteristically eccentric in position (figs. 3, 4, Pl. 32; fig. 13, Pl. 33). The cytoplasm becomes well defined and is often concentrated on opposite sides of the nucleus, forming the cytoplasmic nodules described by several workers. This condition is not constant, the cell often being roughly triangular in shape, due to pressure of surrounding cells (fig. 13, Pl. 33). The cytoplasm is best observed in material fixed with Flemming’s strong solution without bleaching. The fixative stains the cytoplasm a light brown colour and brings out its granular appearance (fig. 16, PL 34) better than any counterstain that was employed. In the full-grown germ-cell (fig. 4, Pl. 32), the nucleus is 7-8 microns in diameter and closely resembles that of the mature oogonia of the adult worm. Fig. 13, Pl. 33, is a photograph of a cross-section through a young redia just behind the gut, showing several of these mature germ-cells. Sections posterior to the one represented in this figure contained varying stages of germinal differentiation as represented in figs. 1, 2, and 3, Pl. 32. As the zone of growth and differentiation of the germ-cells proceeds posteriorly, the fully grown cells may be observed to undergo equal division, resulting in an increase in the number of mature germ-cells which, at a later stage, may form a distinct group (fig. 19, Pl. 34) at the posterior end of the body-cavity.

By repeated division, the mature germ-cell of the redia forms a cercarial embryo or germ-ball without the intervention of germ-masses. The first cleavages were studied with especial care, since they have been reported by several workers to be maturation divisions. Early divisions are characterized by the appearance of a typical somatic spireme stage (fig. 6, Pl. 32). The spireme appears to be a continuous thread, although this observation is based on reconstructions and the finding of exceptionally long pieces in single sections. In no case has this thread appeared double nor has there been observed any condensations that could be interpreted as syndesis or synizesis. The spireme apparently breaks into segments which go directly into the spindle as twelve chromosomes identical in form with those observed in the dividing cells of the mature worm from the vertebrate host. Spindle fibres and distinct centrioles are seen in favourable material. The first cleavage results in the formation of a smaller micromere and a larger macromere. The micromere, which is shown in fig. 5, Pl. 32, in a stage of condensation with two karyosomes, is the next to divide, resulting in a three-cell stage composed of the macromere and the two micromeres. The micromeres divide next, producing a five-cell embryo, three cells of which are shown in fig. 7, Pl. 32. This figure also shows the chromosomes on the equatorial plate of the macromere which is dividing to form a six-cell germ-ball, composed of four micromeres and two macromeres. It has not been definitely determined which of these macromeres divides next, but one of them apparently does and gives rise to a macro-mere and the seventh or investing cell (fig. 8, Pl. 32). Both macromeres, as shown with divided karyosomes in fig. 8, Pl. 32, are the next to divide. Fig. 16, Pl. 34, is a photograph of a later stage, showing the investing cell and four cells produced by the division of the two macromeres. The succeeding divisions of the cells of the germ-ball are exceedingly difficult to follow on account of their irregularity, and the characteristic degeneration of many of the nuclei of the germ-ball. Fig. 10, Pl. 32, which is a section of the later embryo, shows one of these degenerating nuclei and one, possibly two, investing cells. After the first cleavage, cell boundaries of the germ-ball disappear (fig. 16, Pl. 34) and the nuclei appear to be imbedded in a common cytoplasmic mass. The chromosomes shown in microphotographs 17 a and 17 b, Pl. 34, which are consecutive sections of a dividing macromere in a young germ-ball, are represented also in figs. 9 a and 95, Pl. 32. The chromosomes are diploid (twelve) in number and exhibit the same characteristic sizes and shapes that were described for the chromosomes of the adult (Cable, 1931). Since they have often been observed in metaphase, the diploid number of chromosomes in the germ-ball stage is unmistakable. Figs. 11 (Pl. 32), 14,15 (Pl. 33), 16,19 (Pl. 34), 21 and 22 (Pl. 35), which represent sections of older rediae, show various stages of germball development from the undifferentiated primordial germcells to mature cercariae. The embryos towards the anterior end of the rediae are always the more advanced. Between them and the posterior end there is a continuous gradation of development. In younger rediae, undivided and even undifferentiated germ-cells are found clearly distinct from the cells of the wall at the posterior end of the body-cavity. In more mature rediae, which contain mostly advanced cercarial embryos, the movements of these cercariae may change the position of the younger embryos so that the order of development and differentiation is not as apparent as in younger rediae ; but even in these cases, the natural arrangement is rarely so changed that it cannot be recognized. The germ-cells at the posterior end retain their position until the last of them has started to divide and the redia becomes a degenerate sac filled with cercariae.

The exact relation of the body-wall to the germ-cells of the youngest redia observed (fig. 12, Pl. 83) is difficult to determine, due to the lack of differentiation at this stage. As described above, the wall is not well defined and its cells cannot be distinguished cytologically from the germ-cells contained in the vaguely delimited body-cavity or from the cells of the supporting tissue of the sucker. Nuclei are often observed lying directly on the strands of connective tissue marking the inner limit of the body-wall. It cannot be stated with certainty whether these nuclei are germinal or somatic in nature. As soon as growth and differentiation of the germ-cells are initiated, however, the bodywall becomes well defined (figs. 13, 14, Pl. 33), although cell boundaries remain indistinct. Nuclei are fairly numerous in the thick body-wall of the young redia which contains only a few small germ-balls at the most advanced stages of development (fig. 14, Pl. 33). Posterior to the gut, these nuclei are of the lightly-staining type, except for the occasional occurrence of a small dark nucleus similar to those found in large numbers in the heavy supporting tissue of the sucker. As the redia increases in size, the nuclei of the body-wall appear to become less numerous. This observation is not believed to indicate that the nuclei of the wall enter the body-cavity, becoming germcells. Although definite counts have not been made, it seems instead that, due to a lack of multiplication of these nuclei, their number remains fairly constant and, as the redia becomes larger, the nuclei of the body-wall are separated (fig. 22, Pl. 35). When the body-wall becomes greatly expanded, the nuclei are flattened (fig. 21, Pl. -35) much as described for the rediae of several other trematodes. No convincing evidence has been found at any stage that germ-cells arise from the body-wall, although they, as well as germ-balls, may appear in contact with it. This association is believed to be a secondary one.

It approximately the time that germinal development is initiated in the young redia, the cells of the supporting tissue of the sucker become differentiated into the two types mentioned above. The small opaque nuclei appear scattered among the nuclei of the larger type, and also as a distinct cluster (fig. 15, Pl. 33) at the side of the gut. They appear elsewhere in the supporting tissue singly or in small loose clusters (fig. 18, Pl. 34), and one or more may lie very close to a large vesicular nucleus, forming a group which resembles a macromere with micromeres or, due to the dark appearance of the smaller cells, might erroneously be interpreted as an ovum with polar bodies. The large vesicular nuclei with their irregular mass of cytoplasm resemble the parenchymal cells of the adult worm.

Soon after their appearance, the investing cells develop into a distinct membrane completely enclosing the germ-ball. In the more advanced germ-ball, many of these investing cells have become extremely flattened and stain an intense blueblack. The nuclei of the germ-ball retain their germinal appearance for a time; but, as the embryo becomes larger, they decrease in size and are stained more lightly. All stages of this differentiation can be followed continuously from the posterior to the anterior end of the redia. The older germ-balls are characterized by an extremely loose organization (fig. 21, Pl. 35) and, in most preparations, the younger embryos contain distinct and characteristic cavities. Although this condition is enhanced by fixation effects, since it is more apparent in the deeper regions of the infected liver where the fixative did not penetrate properly, it is characteristic of the most favourable material as well. This loose organization is suggestive of dissociation but a careful examination renders this interpretation untenable. There is no evidence that the embryos dissociate completely. The masses of cells which appear to be more or less isolated by the loose organization do not have any features indicating that they may be secondary embryos. As a matter of fact, this dissociationlike appearance is found in embryos possessing well-defined cercarial organ anlagen, the first of which to become differentiated are the large penetration gland-cells and the nuclei of the cells which will form the future excretory bladder. The eyespots and rudiment of the oral sucker are the next well-defined cercarial structures to appear. Germ-cells, recognizable as such, are first seen at a later stage as a number of small opaque nuclei just anterior to the excretory bladder. It has not been possible to trace them back to an early germ-ball stage. Fig. 20, Pl. 35, shows the genital anlage between the excretory bladder and the penetration glands of a mature cercaria outside of the redia. The anlage shows no evidence of a separation into gonads, and its nuclei resemble those of the primordial germ-cells of the young redia. They do not undergo further differentiation until the metacercarial or early adult stage, in which the anlage is separated into the testes, ovary, and accessory structures.

A considerable number of nuclei in each germ-ball seem to degenerate, if their reduced size and intense staining can be taken as an indication of that condition. For convenience in later discussion, these structures are designated as ‘degenerating nuclei’ although their exact nature is as yet undetermined. These small opaque nuclei, which appear during early cleavage, resemble extranuclear karyosomes (fig. 11, Pl. 32; figs. 16, 19, Pl. 34). They are first recognized by a decrease in size and an accompanying condensation of chromatin (fig. 10, Pl. 32). In later stages of the germ-balls (fig. 22, Pl. 35), they are found in greater numbers and appear in Crytocotyle much as figured by Looss (1892) forAmphistomum subclavatum. They have probably been regarded by some investigators as degenerating polar bodies but it seems that they are too abundant to be interpreted in that manner. Furthermore, entire germ-balls are commonly observed in a state of degeneration (fig. 19, Pl. 34; fig. 22, Pl. 35). These degenerating germ-balls, which were also observed by Looss (1892), Tennent (1906), and possibly others, are usually seen near the posterior end of the redia as a mass of darkly-staining nuclei, varying in size. Macromeres and micromeres can be recognized in some cases and there are often two or more such masses in a redia. These degenerating germ-balls eventually disintegrate and their components are observed scattered among the cercarial embryos where they might easily be mistaken for primordial germ-cells or interpreted as polar bodies by one not familiar -with the material. Their degeneration products may serve to nourish the remaining germ-balls. It will be recalled that in the adult (Cable, 1931) degenerating nuclei were observed in the cytoplasm of the oocyte just before fertilization.

In considering the evidence concerning germinal lineage in the digenetic trematodes, it is patent that studies dealing only with sporocyst and redial generations can render conclusions that are applicable to those stages alone. The concept of germinal lineage assumes that by the division of a stem-cell early in ontogeny, the soma and germ become separated into two distinct lines. The tissues of the somatic line enclose, nourish, and protect the cells of the germinal line. The early segregation of germ plasm, and its subsequent continuity and cytological distinction from the soma, are well demonstrated in several forms including Ascaris, certain of the coelenterates, and many insects and crustaceans. With the exception of the hydromedusae, the life-cycles of these forms are relatively simple as compared with those of the digenetic trematodes. In the latter group, where three or even more hosts may be necessary for the completion of the life-cycle, the germinal line may not be as distinct throughout the cycle as has been supposed from studies dealing with only the larval stages in the molluscan host. Many investigators, furthermore, have described the origin of germcells from the body-wall of sporocysts and rediae, and hence maintain that the germinal line is discontinuous in these stages. Only the definite demonstration of the production of germ-cells from the soma of sporocysts, rediae, and cercariae can disprove the concept of germinal continuity in the Digenea. If the cells of the germinal line are segregated early in ontogeny and remain distinct from the soma throughout the life-cycle, the requiremerits for germinal lineage are fulfilled regardless of whether polyembryony, parthenogenesis, or hermaphroditism is characteristic of multiplication during larval stages.

Leuckart (1886) suggested that the germ-cells of sporocysts and rediae are derived directly from the fertilized egg of the adult and are merely enclosed by the soma of the larval stages. Kathariner (1904,1920), Dollfus (1919), and Brooks (1928,1930) have been the chief advocates of this hypothesis and all have postulated germinal continuity from fertilized egg to fertilized egg, although their studies have not dealt with entire life-cycles. Several investigators including Looss (1892), Tennent (1906), Faust (1917), and Dubois (1929) have described the origin of germ-cells from the body-wall of sporocysts and rediae. These observations oppose the concept of germinal continuity since they indicate that the germ-cells of larval stages are not directly descended from the fertilized ovum. A consideration of the development and differentiation of cercariae also throws doubt on the continuity of the germinal line in that stage of the life-cycle.

In Cryptocotyle, the germ-cells are not recognizable until the germ-balls contain several hundred cells and the penetration glands, excretory vesicle, and the oral sucker have become differentiated. Descriptions of other species of cercariae indicate a great variation in the degree of sexual development of the ‘mature’ larvae. In a few species, the descriptions definitely state that the genital anlagen were not observed, while a number of workers have not mentioned the germ-cells at all. In view of the fact that the mature cercaria is a highly differentiated larva, it seems that at least recognizable germ-cells would be present if germinal lineage continued through this stage of the life-cycle. In other species of cercariae, more or less well-developed genital anlagen are present in mature specimens. In Cercaria wardi, Cercaria inversa, and Cercaria douthitti, the gonads are represented by a single undeveloped mass of germinal cells. Distinct testicular and ovarian anlagen with lines of nuclei representing the gonoducts are described for Cercaria inhabilis, Cercaria robusta, Cercaria infracaudata, and many others, while Cercaria fusca and Cercaria macrostoma are sexually mature with eggs in the uterus at the time the larvae leave the snail. This great variation in sexual differentiation of mature cercariae may be due to:

1. An earlier segregation of germ-plasm in some species than in others.

2. Longer or shorter periods of development in the mollusc.

3. More rapid germinal development in the case of larvae having well-defined gonads and gravid uteri.

There is evidence for germinal continuity, however, through a part of the life-cycle. In general, the miracidia of trematodes contain indisputable germ-cells at an early stage and a few enclose well-formed rediae at the time of hatching. Since the miracidium metamorphoses into the mother-sporocyst and the germ-cells are undeniably carried over in the body-cavity of the latter stage, germinal lineage must be accepted for that phase of the life-history in many trematodes. This early segregation of germ-cells in the miracidium has been well established for so many species, that, in the absence of direct observation of the miracidium-mother-sporocyst, it is believed that Cryptocotyle lingua is probably no exception to this rule. It cannot be stated at this time whether or not somatic cells of the more mature sporocyst contribute to the reproductive cells of that stage. It is not believed likely, however, in view of observations on germinal development in the redia. The germ-line continues through the redial stage of Cryptocotyle where it is represented by the small primordial germ-cells of the very young redia. As described above, these germinal cells are so similar to all other cells of the young redia that they can be distinguished with certainty only by their position, and even this criterion is difficult to apply in the case of some nuclei that appear to lie very close to the faintly delimited inner boundary of the body-wall. Cells of the wall have never been observed to exhibit any characteristics indicating that they contribute to the germinal line. It is concluded, therefore, that germinal lineage occurs with certainty in the redia and probably also in the sporocyst since it is still nearer the fertilized egg than is the redial generation.

Germinal lineage in Cryptocotyle seems to disappear, however, after the initiation of germ-ball formation since germcells are not recognized in the cercarial embryo until a late stage of development. It seems that at least some of the cells of the germ-ball are equipotent or totipotent in the sense that they are not restricted to either somatic or germinal lines, but give rise, at a later stage, to elements of both. It might be argued, on the basis of Brooks’ (1930) hypothesis, that germinal precocity, which serves to continue the germinal line in sporocysts and rediae, becomes so decreased in intensity in the germball that somatic factors dominate development. As a result, the germ-plasm becomes inhibited to such an extent that its lineage is interrupted and the germ-cells arise at a later date from equipotent cells that are, until definite segregation, as much somatic as germinal in nature. One would expect an earlier segregation of germinal elements than has been observed in the cercariae of Cryptocotyle and many other species, if germinal continuity is as distinct as Kathariner (1920), Dollfus (1919), and Brooks (1930) have maintained. After the appearance of the genital anlage in the cercaria, the germinal line becomes re-established and continues in a slowly differentiating state through the metacercarial stage. After the parasite becomes established in the final host, germinal development is accelerated and sexual maturity is quickly attained.

It should be stated again that no evidence has been found in the present study that would indicate the occurrence of maturation or other sexual phenomena in the redial stage of Cryptocotyle. According to Brooks (1930), the following may be misinterpreted as polar bodies :

‘1. Small darkly-staining cells such as were described by Reuss (1903) as polar bodies but shown by Rossbach (1906) to be of another nature.

‘2. Micromeres embedded in cytoplasmic nodules as are described by Faust (1917).

‘3. Karyosomes of cells lying below or above the centre of focus, or belonging to cells in which the remainder of the cell did not take the stain well.

‘4. Snail eggs undergoing maturation. This error is very easily made by workers new to the material.

‘5. Parts of karyorrhectic nuclei.

‘6. Artifacts such as dirt, undissolved stain, &c.’

Other things that may be mistaken for polar bodies are :

1. The components of degenerating germ-balls which have been described by a number of workers and also frequently observed in Cryptocotyle.

2. The association of one or more small dense nuclei with a larger vesicular nucleus in the body-wall and supporting tissue of the sucker. Brooks (1930) has pointed out that the sporocysts of some trematodes have thickened portions of the body-wall which are in reality rudimentary suckers. Such structures with parenchymal cells similar to those of Cryptocotyle could be easily mistaken for ovaries containing maturation stages.

The micromeres of the young germ-balls are the most likely of all structures to be misinterpreted as polar bodies. The writer’s study of the chromosomes of Cryptocotyle during early division stages precludes any possibility of reduction. The diploid chromosome number persists from the fertilized ovum of the adult, throughout larval development and until the first maturation division of the oocytes and spermatocytes in the next adult generation. It might be argued that pseudomaturation, which is characteristic of diploid parthenogenesis in such forms as copepods, echinoderms, and certain insects, occurs in the trematodes and that the micromeres are true polar bodies. The absence of any meiotic-like phenomena such as pseudoreduction, synapsis, syndesis, &c., negates the occurrence of pseudomaturation in Cryptocotyle rediae. On account of their number, it cannot be maintained that the degenerating nuclei of the germ-ball are polar bodies. It is concluded, therefore, that no form of maturation occurs in the redial stage of Cryptocotyle lingua.

Considerable emphasis has been placed on the presence or absence of ovaries in sporocysts and rediae by a number of investigators. Since true ova do not occur in the rediae of Cryptocotyle, it is apparent ipso facto that a true ovary would not be present. Germinal multiplication does take place by equal division of mature germinal cells before cleavage begins. In many preparations, a mass of mature cells, none of which had started cleavage, was observed at the posterior end of the redia (fig. 19, Pl. 34). There is no evidence that such a group is derived from a single cell in the manner described by Brooks for other trematodes ; the mass cannot therefore be interpreted as a germ-mass in the sense implied by him. The region in which this germinal multiplication by equal division occurs at a given time may be termed the zone of multiplication. Behind this zone in the young rediae, all stages of growth and differentiation of germinal cells were observed. In older rediae, however, the zone of multiplication moves posteriorly until all the germinal cells are finally included in it.

It has already been mentioned that, in the very young redia, cell boundaries are indistinct, the body-cavity is poorly defined and a considerable amount of parenchymal tissue is present. It is believed that the strands of this tissue may play an important part in the interpretation of germinal development. As the first of the germinal cells present in the young rediae become mature and begin to divide, considerable pressure must be exerted on the surrounding cells, and the abundant strands of connective tissue may be pressed forward and backward with the result that the growing and dividing germinal cells appear to be enclosed in a membrane. As development of the enclosed cells ensues, the strands surrounding the group are parted and the contained embryos are freed into the body-cavity. The bodycavity itself seems to become well defined only after germinal development has broken down the obscuring net of connective tissue.

Fig. 22, Pl. 35, is a photograph of a redia showing groups of embryos suggestive of the dissociating germ-masses described by Brooks (1930). A careful study of this and other similar sections revealed that, when present at all, the restraining sheath of connective tissue was found only on the side next to the body-cavity. It was found also that, ranging from the more advanced embryos near the open body cavity, there is a gradation posteriorly to undivided and even undifferentiated germinal cells without a continuous intervening membrane such as would be expected if the group of cells, similar to those seen in fig. 22, Pl. 35, constituted a true germ-mass in the sense of Brooks. The section represented in this figure is not a longitudinal section through the posterior end of the redia since this part of the larva was bent at a considerable angle to the major portion of the body, the angle at which the section was cut being indicated by the apparent thickness of the body-wall at the end of this section. It is evident, then, that serial sections must be studied with especial care in obtaining an accurate picture of the redia as a whole. Aside from such observations as the one just explained, no phenomena that could possibly be interpreted as germ-mass formation and dissociation have been observed in the rediae of Cryptocotyle. The loose organization of cercarial embryos has been shown in the preceding section to be not a true dissociation but a regularly occurring feature of cercarial development. Although Brooks did not observe germmasses in some of the species of cercariae studied by him, he concluded that they were formed during larval multiplication, since he did find ‘ex-components’. It seems quite possible that these cells may have developed in a more direct manner similar to that observed in Cryptocotyle.

In attempting to determine the significance of the apparent degeneration of germ-balls in the rediae of Cryptocotyle, it was found that a possible explanation may be afforded by the mesozoan, Dicyema. Although the writer has not studied the Mesozoa (Rhombozoa) at first hand, the account of Hartmann (1925) makes it possible to compare certain features of the life-history and germinal development of Dicyema with those of the trematodes. In the young agametic generation of this mesozoan, the centre of the body is occupied by a single, multinucleate axial cell. The nuclei of this cell are of two types, a single primary vegetative nucleus and a number of germinal nuclei, both types arising from division of the primary nucleus of the axial cell. The primary vegetative nucleus divides, giving rise to a secondary vegetative nucleus. The degenerating germballs of the trematodes, like the vegetative nuclei of Dicyema, are also derived from cells similar to and very closely associated with the germinal cells which produce normal embryos and, if they have any function at all, it may be that of nourishing the surrounding embryos. These considerations suggest that the degenerating germ-balls may be descendants of a cell in the young redia which corresponds to the vegetative nucleus of Dicyema. Certain observations indicate that this may be the case. In fig. 14, Pl. 33, a longitudinal section of a young redia, two large and two small darkly-staining nuclei are seen between the zone of mature germ-cells and that of the intermediately differentiated germ-cells. Since these nuclei, and also the degenerating germ-balls at a later stage, lie considerably anteriorly to the primordial germ-cells, and are separated from them by a number of intermediate stages of differentiation (in the young redia) and cleavage (in the older redia), it is believed that these nuclei give rise to the embryos which degenerate after reaching a considerable size. It seems quite possible, in view of these observations, that these dense nuclei correspond to the vegetative nuclei of Dicyema. Proof of this relationship is lacking since the writer has not studied Dicyema from actual material, but, if it is true, it affords additional evidence that the Mesozoa are degenerate flatworms and not forms intermediate between the Protozoa and the Metazoa. In the cleavage of both the agamete of Dicyema and the mature germinal cell in the redia of Cryptocotyle, a macromere and a micromere are produced by the first division and the five-cell stage consists of a macromere surrounded by four micromeres. The condensation of cytoplasm around the differentiating germcell of Dicyema also has a counterpart in the differentiation of the primordial germinal cells in the redia of Cryptocotyle.

Charts which illustrate in graphic form the various hypotheses of trematode germinal development have been helpful in evaluating the observations of other investigators and interpreting the germ-cell cycle of Cryptocotyle. The great variation in the life-histories and in the details of development, that have been reported, render it impracticable to construct charts showing all of these differences. An attempt has been made, therefore, to combine these variations in four diagrams representing the principal theories of germinal development in the Digenea. These charts are intended not to represent the number of cell divisions, but rather to show only the general features of germinal development and its relation to the soma. Since metamorphosis extending over several generations does not account for germinal multiplication in larval stages, it may be regarded as applicable only to portions of the life-cycle and thus may be applied to any of the four hypotheses discussed.

Text-fig. 1 represents the concept of metagenesis or alternation of generations. This theory implies the alternation of sexual and asexual phases of reproduction and interprets the production of one or more intercalary stages of the trematode life-cycle as a process of internal budding from the wall of the sporocyst or redia. The cells giving rise to daughter-sporocysts or rediae may have a number of possible origins : they may come directly from the fertilized egg as the only source ; this direct origin may be supplemented by budding from the wall of the mothersporocyst; or they may be produced only from cells of the sporocyst wall that remain equipotent for germ and soma until the development of daughter embryos is initiated. The latter possibility is believed unlikely in view of the great mass of evidence indicating early segregation of germ-cells in miracidia. The stage of asexual budding, necessary for true metagenesis, may conceivably be represented, however, in the daughtersporocyst or redia, in which a number of investigators have reported that cercarial embryos develop from cells budded off from the body-wall. Such a process indicates, as shown in the chart, that the germ-cell of the mother-sporocyst gives rise (1) to cells that are purely somatic in nature and compose the sucker, supporting tissues, &c., of the redia; and (2) to cells that, while somatic in appearance and composing the body-wall, have the potentiality of budding or passing directly into the body-cavity and giving rise to cercariae. This process would involve an interruption of germinal lineage in the daughtersporocyst or redia. This discontinuity is represented in the chart by half-shaded circles which indicate cells equipotent for germinal and somatic elements. Development of post-redial stages is also represented by equipotent cells whose potency is not restricted until the primordial germ-cells of the cercaria are segregated.

Text-fig. 2 represents the concept of heterogeny which postulates the alternation of parthenogenetic and sexual generations in the trematode life-cycle. The validity of this hyopthesis depends on the demonstration of definite maturation phenomena in the germ-cells giving rise to daughter-sporocysts, rediae, and cercariae, since it must be shown conclusively that these cells are true ova. It should be pointed out that the concept of heterogeny does not necessarily imply discontinuity of the germinal line, since maturation phenomena could be associated with reproductive cells that are segregated early in ontogeny. Most of the advocates of this hypothesis, however, have reported that parthenogenetic ova are somatic in origin or at least are derived from the body-wall. A definite ovary has been described in some species. Three possible origins of parthenogenetic ova in the redia are represented in the chart :

1. From cells of the body-wall. Described by Tennent (1906) and Faust (1917) for certain species. This mode of origin invalidates germinal lineage.

2. Segregation of germ-cells in the young larvae and also production of parthenogenetic ova from the body-wall. Implied by Cary (1909).

3. Early segregation of germ-cells alone. Not definitely postulated by any advocates of heterogeny. Possibly true of the ‘ovaries’ and ‘specialized germinal masses’ described by Haswell (1903) and Faust (1917).

In Text-fig. 2, each parthenogenetic ovum is represented as producing a single polar body; as many as three have been reported.

Brooks’ (1930) hypothesis of germinal lineage with polyembryony is represented in Text-fig. 3. Germinal lineage is continuous from fertilized ovum to fertilized ovum and the cells of the germinal line are regarded as merely enclosed by the soma of the sporocyst and the redial stages which are themselves sterile. Brooks interprets as polyembryony the formation and dissociation of primary, and perhaps secondary and tertiary germ-masses whose components give rise to the next stage of the life-cycle. In this text-figure, a single germ-mass is represented for each generation involved in larval multiplication. It has been mentioned in the discussion of germinal continuity of Cryptocotyle that continued lineage through post-redial stages, as postulated by Kathariner (1920), Dollfus (1919), and Brooks (1930), is not supported by observations that, may be regarded as conclusive in nature.

Text-fig. 4 represents Woodhead’s interpretation of the germinal line in Bucephalus as polymorphism with three sexually mature hermaphroditic generations. Although Woodhead did not postulate germinal continuity by early segregation of germ-cells in the various stages, for the sake of simplicity the writer has taken the liberty to represent this condition in the chart. The fertilized ovum of the final adult gives rise to a branching sporocyst or ‘dendritic colony’ whose several tips possess functional ovaries and testes. Gametogenesis and fertilization occur in this stage in a manner similar to that in the final adult worm. The fertilized ovum of the sporocyst gives rise to a ciliated pro-redia which metamorphoses into an hermaphroditic redia also possessing functional gonads. The fertilized ova of the redia produce cercariae which develop into the fina hermaphroditic adults. In a recent note (1932), Woodhead has reported that Leucochloridium has a germinal cycle similar to that of Bucephalus. This observation makes it impossible to regard polymorphism as restricted to the Bucephalidae, although this group is a peculiar one in many respects and may not be closely related to other digenetic trematodes.

The impossibility of explaining Woodhead’s observations on Bucephalus and Leucochloridium on the basis of Brooks’ concept of polyembryony makes it necessary to approach the problem of trematode germ-cell cycles from an altogether different angle. It appears certain that the only satisfactory interpretation must be based primarily on the phylogeny of the Digenea. It must take into consideration all well-substantiated findings concerning germinal development in the members of this group. It must regard the Digenea as descendants of free-living ancestors and must determine the manner and extent of modification of the germ-cell cycle as a result of the parasitic habit. Although evidence is as yet extremely fragmentary and there is considerable disagreement in the observations and interpretations of different workers, sufficient data are available to suggest a phylogenetic hypothesis of trematode germinal cycles. Woodhead’s observations support the view that all digenetic trematodes were primitively parasites of molluscs, becoming sexually mature and completing the lifecycle in those hosts. The two genera, Bucephalus and Leucochloridium, in which Woodhead has reported hermaphroditic stages in the mollusc, are quite dissimilar. It is of interest to note that in both, until quite recently, cercariae have been regarded as arising directly from the mother-sporocyst. The redia of Bucephalus, according to Woodhead (1931) is inconspicuous, and its primitive nature is indicated by the ciliated covering of the pro-redial stage. Bucephalus and Leucochloridium have the minimum number of intercalary stages in the life-cycle and, if sexual maturity characterized the primitive trematode in the mollusc, these forms may be regarded as having been changed to a less extent by parasitism than most other trematodes. A ciliated pro-redia, such as that described by Woodhead for Bucephalus, has not been reported for any other trematode, unless Sewell (1922) was mistaken in his observations on Cercaria indicae XV in which he noted that the sporocysts gave rise to both cercarial embryos and miracidia. These miracidia resemble the prorediae of Bucephalus in that they are encapsulated, ciliated, and arise from sporocysts. If Sewell’s interpretation is correct, it affords additional evidence that the primitive trematode lifecycle was direct and completed in the mollusc. The occurrence of sexually mature metacercariae, as described by Dollfus (1929, 1932), however, probably cannot be interpreted as an intermediate stage in the evolution of the digenetic life-cycle. It is explained perhaps more satisfactorily by Dollfus as progenesis with extreme sexual precocity and the possibility of both an abbreviated and an extended life-cycle.

Brooks’ hypothesis argues for a sudden change in the lifecycle from sexual reproduction to polyembryony on the adoption of parasitism. His concept is an exceptionally fortunate one in many respects since it can explain the presence or absence of germ-masses and any number of intercalary stages. This hypothesis could be easily applied to Cryptocotyle by assuming that germinal precocity is less marked in this species than in those studied by Brooks, and that the germ-cells, instead of forming germ-masses, have a ‘normal’ tendency to produce larvae directly. Woodhead’s findings, however, cannot be explained on the basis of Brooks’ hypothesis, adaptive as it is.

It is believed, therefore, that changes due to parasitism have appeared in a gradual rather than a sudden manner, and that the primitive trematode became sexually mature in the mollusc and completed the life-cycle in that host.. With the evolution of parasitism, sexual reproduction in the mollusc has gradually disappeared and accessory stages have been acquired. If this is true, it would not be surprising, but indeed expected, that only the male reproductive processes have been lost in some of the present trematodes and their development is essentially parthenogenetic. In perhaps the majority of digenetic trematodes, the life-cycle has been modified to a still greater extent with the result that sexual phenomena in the molluscan host have entirely disappeared and polyembryony, with or without distinct germmasses, is operative in larval multiplication. The life-cycle in the mollusc may thus have been considerably modified before the vertebrate host was parasitized. After the appearance of vertebrates, the trematodes became established in that group for some reason as yet unknown. The inclusion of the vertebrate host in the life-cycle may have been incidental to its feeding on infected molluscs; or it may possibly have been the result of changes in the trematode as a result of parasitism of the mollusc such that further development in another host became necessary for survival. The concept of gradual change is admittedly hypothetical because of insufficient information, but it seems on the whole more reasonable and applicable to all the facts than the hypothesis that adoption of the endoparasitic habit resulted in sudden changes in germinal development similar to those that have been induced experimentally in other forms by such exogenous factors as lowered oxygen tension, temperature effects, &c. It cannot be denied that such factors have modified the lifecycle in the trematodes, but, in view of the findings of various workers, it seems unlikely that they have induced sudden intercalations of polyembryonie phases in the germ-cell cycle.

The germ-cell cycle of Cryptocotyle is represented in Text-fig. 5. It is concluded that this species is intermediate in position in the proposed phylogenetic scheme of germinal development, since the cercariae develop directly without the intervention of germ-masses in the sense of Brooks, since maturation or other sexual phenomena are lacking in the redia, and since the life-cycle does not have daughter-sporocysts or daughter-rediae. Patterson (1927) defines polyembryony as the production of two or more individuals ‘from a single fertilized egg during the course of its early development’. An examination of the literature shows that this phenomenon is extremely variable in nature. In some forms, the first divisions of the fertilized egg may definitely separate and initiate the development of two or more embryos, while in one notable case, that of the armadillo, the fertilized ovum gives rise to a relatively undifferentiated structure, the blastocyst, which shows no indication of producing four instead of a single young until it reaches a considerable size. On account of such cases of delayed embryo formation, a definite distinction between metagenesis and polyembryony does not exist. Stockard (1921), indeed, has pointed out that the embryonic mass, giving rise by budding to a number of embryos, may be regarded as the sexually produced generation. Reproduction in this generation by budding may therefore be regarded as asexual in nature and the lifecycle may be interpreted as metagenetic.

In view of the early segregation of germinal cells in the rediae of Cryptocotyle, and since the cells of this line multiply by equal division, larval reproduction in this trematode is interpreted as a type of polyembryony. The apparent interruption of germinal lineage in the germ-ball, as represented in Text-fig. 5, is coincident with the development of such specialized cercarial structures as the tail and larval glands which are necessary for establishing infection in the fish host. Concomitant with this somatic differentiation, germinal development is repressed and the genital anlage does not become apparent until a rather late stage. After the parasite becomes established in the final host, germinal development proceeds rapidly and gametogenesis, fertilization, and cleavage follow as represented in the chart and previously described in detail (Cable, 1931).

A consideration of the diverse life-histories and of the phylogenetic affinities of the Digenea, in addition to specific investigations of germinal development that have been made, suggests that any single type of germ-cell cycle may not characterize all members of the order. The writer, therefore, proposes the above phylogenetic concept as a basis for further study rather than a compromise between existing divergent theories. Before an understanding of germinal development in the Digenea as a group will be possible, more data must be secured from a large number of complete life-histories in which the germ-cell cycle is followed from adult to adult. As more life-histories become known, a comparative study will be made possible, and many of the problems related to the origin of complicated life-histories in the Digenea will be better understood.

ct, connective tissue; dgb, degenerating germ-ball; dn, degenerating nucleus of germ-ball; ga, genital anlage; gb, germ-ball; gc, germ-cell; gl, gut ; ic, investing cell ; igc, immature germ-cell ; k, karyosome ; mac, macromere ; mgc, mature germ-cell ; mic, micromere ; pgc, primordial germcell; s, sucker.

Fig. 1.—Primordial germ-cell of the young redia (× 2,466).

Figs. 2, 3, and 4.—Progressive stages in differentiation of germ-cells of the young redia ( × 2,466).

Fig. 5.—Two-cell stage of the germ-ball ( × 2,466).

Fig. 6.—Spireme stage typical of cell divisions in the germ-ball ( × 2,466).

Fig. 7.—Three cells of the five-cell embryo showing division of the macro-mere to form the six-cell stage ( × 2,466).

Fig. 8.—Seven-cell stage with investing cell ( × 2,466).

Fig. 9 a and 6.—Twelve chromosomes in the dividing macromere of a young germ-ball as observed in two consecutive sections and photographed in fig. 17 a and b ( × 2,466).

Fig. 10.—Section of a later stage showing the investing cell and a degenerating nucleus ( × 2,466).

Fig. 11.—Camera lucida drawing of a young redia showing various stages of germ-ball development ( × 466).

Fig. 12.—Longitudinal section of a very young redia (microphotograph, × 575).

Fig. 13.—Cross-section at a level immediately posterior to the gut of a young redia showing mature germinal cells with typically eccentric karyosomes (microphotograph, × 1,000).

Figs. 14 and 15.—Longitudinal sections of young rediae showing differentiating germ-cells and young germ-balls (microphotographs, × 500 and × 600 respectively).

Fig. 16.—Section through the posterior end of a redia showing germ-balls with investing cells and degenerating nuclei (microphotograph, × 1,100).

Fig. 17 a and 6.—Consecutive sections of a young germ-ball showing the chromosomes of a dividing macromere (microphotograph, × 1,250).

Fig. 18.—Cross-section through the anterior end of a mature redia showing the oral sucker and the nuclei of its supporting tissue (microphotograph, × 750).

Fig. 19.—Longitudinal section through the posterior end of an older redia (microphotograph, × 900).

Fig. 20.—Frontal section of a mature cercaria (microphotograph, × 625).

Fig. 21.—Longitudinal section through an older redia showing undivided germ-cells at the extreme posterior end and the loose organization of cercarial embryos (microphotograph, × 455).

Fig. 22.—Section through the posterior end of a redia showing various division stages and a degenerating germ-ball (microphotograph, × 500).