Some remarks on North Atlantic Non-Pelagic Polystylifera. Dr. Gerarda Stiasny-Wijnhoff.

Transcription

1 Some remarks on North Atlantic Non-Pelagic Polystylifera. By Dr. Gerarda Stiasny-Wijnhoff. Leiden. W i t h Plates C O N T E N T S. p A G E 1. I N T R O D U C T I O N P U N N E T T I A S P L E N D I D A Keferstein U N I P O R U S B O E E A L I S P u n n e t t HtTBRECHTONEMERTES LANKESTERI H u b r e c h t E X P L A N A T I O N O F P L A T E S INTRODUCTION. THIS publication contains the description of three previously but insufficiently known representatives of the former genus Drepanophorus, that live in the northern Atlantic Ocean. They are, together with the two species of Uniporus described by Brinkmann (1913), the only Reptantia that are known from the Atlantic. One of them, Punnettia splendida, lives in the English Channel, and was described by Keferstein and Mclntosh as a distinct species, different from Hubrecht's Drepanophorus rubrostriatus. Burger included all these species in the Mediterranean D. spectabilis, and so did Eiches and the author. After my publication on Mediterranean Polystilifera (1926), in which the Atlantic species as being separate was not mentioned, the late Professor Mclntosh asked why I had been silent on this question. The present article gives my answer. The material has partly been collected by me in Plymouth, and partly consists in a series of sections from Professor Brinkmann. The description of the second species was necessary because Brinkmann had supposed Punnett's Drepanophorus borealis to belong to his new genus Uniporus. As the NO. 306 N

2 168 GERARDA STIASNY-WIJNHOFF original material is in my hands it seemed my duty to give some new information that throws light on the systematic position of this genus. It is in accordance with Brinkmann's views. And thirdly I had the opportunity of studying Huhrecht's sections of Drepanophorus lankesteri, collected by the ' Challenger' Expedition. Several authors have expressed their wish for more detailed information about the structure of this deep-water nemertean. Though the sections are not as instructive as one might wish, I thought it worth while to give all the information I could get from them. It shows that Drepanophorus lankesteri is not the interesting nemertean some authors supposed it to be, and has no nearer affinity to the genus Uniporus. The sections contain only the precerebral, gastric, and tail-region of the animal; there are no sections of the middle part. This might have some influence on my conclusions about the structure of the rhynchocoelomic diverticula as in some cases the branching of these pouches can only be found in the beginning of the enteric region; however, whenever branching was observed, the first rhynchocoelomic pouches lay on the brain, which they here do not reach. The sections therefore suffice to settle this question. Though the rhynchocoelomic diverticula are not branched, and many features are in accordance with our knowledge of the structure of many Mediterranean species, still other features remain that make it necessary to create a new genus for this deep-sea nemertean of the North American waters. I call it Hubrechtonemertes lankesteri in honour of rny deceased teacher, Professor Dr. A. A. W. Hubrecht, who gave the first description of this specimen. PUNNBTTIA SPLENDIDA Keferstein. Figs. 1-5, PI. 11. Keferstein, W. (1862). 'Z. wisa. Zool.', v. 12, p. 59. Borlasia splendida. Diesing (1863). 'Sitzb. Akad. Wien', v. 46, i. p Ptychodes splendida. Mclntosh, W. C. (1869). 'Tr. R. Soc Edinburgh', v. 25, pp. 342, 355. Cerebratulus spectabilis. (1873 4). "BritishNemerteans", 'Ray Soc', Amphiporus speotabilis. Barrois (1877). 'Ann. Soi. Nat.', ser. 6, v. 6, no. 3. Amphiporus spectabilis. Joubin, I. (1890). 'Arch. Zool. Exper.', ser. 2, v. 8. Amphiporus spectabilis.

3 NORTH ATLANTIC POLYSTYLIFEHA 169 Riches (1893). 'J. Mar. Biol. Assoc.', N.S., v. 3. Drepanophorus rubrostriatus. Wijnhoff, G. (1912). Ibid., v. 9. Drepanophorus spectabilis. Habitus. A flat nemertean with thin margins and an oar-like tail. The head is well separated from the body and pigmented. The dorsum has five white lines on a brown field, and white margins. Mclntosh gives a good picture in his monograph (1873, PI. iii, fig. 7). The mouth is separated from the proboscis-pore and lies in its neighbourhood. About seventy eyes are situated in the top of the snout, arranged in four stripes (Keferstein, PI. v, fig. 20). The cephalic slits are large with seven or eight longitudinal secondary slits (Keferstein, PL v, fig. 12). The cerebral pore opens into the outer lateral secondary slit (fig. 1, PL 11). The cerebral organs lie at the side of the brain (Keferstein, PL v, fig. 10) mm. long, 4 mm. broad. Anatomy. The epithelium of the body-wall is rather high, and rests on a well-developed basement-membrane in the pre-enteric region. The circular musculature constitutes a much thinner layer, but in the enteric region it is broader than the basement-membrane. The inner longitudinal musculature is very thick (figs. 1, 3, 4, PL 11) in the cerebral and stomodaeal regions; in the enteric region it is very much reduced, and does not exceed very much the thickness of the epithelium (fig. 5, PL 11). The inner circular muscle-layer is represented by strong dorso-ventral muscles in the brain-region (fig. 1, PL 11), and also by circular muscles around the gastric cavity (fig. 4, PL 11). The body parenchyma is well developed too. The head-region contains but very little parenchyma. The thick epithelium rests on a thin basement-membrane. The whole space inside the thin outer circular musculature is filled by longitudinal musclefibres, a well-developed head-gland, the large eyes, bloodvessels, rhynchodaeum, and oesophagus. Though the inner longitudinal muscle-layer is but very thin (dorsally and laterally) at both sides of the proboscis-pore and the oesophagus a thick longitudinal musculature is developed in the tip of the snout. Here are inserted the muscle-fibres that constitute the mm. retractores capitis. The eyes are found only above these ventral

4 170 GERARDA STIASNY-WIJNHOFF thickenings of the longitudinal musculature. The mouth lies just behind the proboscis-pore, but is quite separate from it. The oesophagus is narrow (fig. 1, PI. 11) and has its own longitudinal muscle-coat. The rhynchodaeum is surrounded by a parenchyma that is devoid of muscle-fibres and has a glandular epithelial wall. Circular and diagonal muscle-fibres converge at a certain distance from this tube in the parenchyma, so as to form a ring, of which the dorsal part is composed of longitudinal muscles. Inside this ring no organs of the head are to be found. The m. constrictor rhynchodaei is in contact with the ventral thickenings of the inner longitudinal musculature, and lies just before the precerebral septum. This has not been dissolved into separate muscles, but forms a true muscular septum. The proboscis shows the usual structure of its wall, with twenty-four to twenty-six nerves. The two parts of the proboscis are outwardly not separated by a constriction; a separate bulbus seems not to be present, nor is there a septum. The base of the stylet is much curved and wears many small stylets. The number of sacs is but small, I counted ten in sections ; each sac contains at least eight small very sharp stylets. The base is situated on a lateral papilla, which nearly fills the bottom part of the proximal tube and leaves only a narrow space round the papilla free. Through this space the proximal and distal parts of the tube communicate; a separate canal piercing through the septum, as for instance in the Monostylifera, does not exist. The epithelium of both parts shows the usual differences. Both kinds of gland-cells are separated by the papilla at one side whereas on the other side the change of glands is a sudden one. The construction of the muscular wall of the proximal part changes just before the region of the papilla. The rhynchocoelomic wall has the usual basket-like structure and is rather thick (figs. 3 and 4, PI. 11). An outer layer of circular muscle-fibres, often formed by the inner circular musclelayer of the body-wall, is absent. The first diverticula are found at the side of the glandular processes of the cerebral organs; they are short and have a dorsal position; later diverticula reach into the margins of the body, and even may be found underneath the nerve-cords (fig. 5, PI. 11). They are unbranched

5 NORTH ATLANTIC POLYSTYLIFERA 171 and thin-walled, but show a thickening of the wall when contracted (figs. 4 and 5, PI. 11). The central nervous system consists of a pair of large brain-lobes, that is connected by a high dorsal commissure, though the proboscidian sheath is not at all enlarged. The dorsal ganglia are very large (figs. 1 and 2, PI. 11) and attain one and a half times the length of the much smaller ventral lobes. The fibrous mass shows traces of bifurcation. A pair of large subdorsal nerves is conspicuous (fig. 4, PI. 11). The ventral lobes are separated from the dorsal ganglia in their distal part and continue laterally into the nerve-cords. These are very large in the gastric region (fig. 3, PL 11) and much smaller in the enteric region (fig. 5, PI. 11). They never touch the bodymusculature and always are surrounded by the parenchyma. Though their position is rather lateral in the gastric and enteric regions (figs. 3, 4, 5, PI. 11), they near each other in the gonadial region leaving about half the breadth of the animal between them. The sense organs consist of a pair of large cerebral organs, situated at the side of the brain, the cephalic slits with seven or eight secondary longitudinal furrows and the eyes. The eyes are arranged in four stripes that do not reach to the brain and are ±70 in number (Keferstein, PI. v, figs. 10,12). The pore of the cerebral canal (fig. 1, PI. 11) opens into the outer of the longitudinal slits and is almost lateral in position (figs. 1 and 2, PI. 11). The cerebral organ is shorter than the ventral ganglia; the short canal bifurcates after its entrance into the organ. The dorsal branch becomes the sac, the ventral one opens into the glandular-tube which continues behind the organ independently (Keferstein, PI. v, fig. 4) (fig. 2, PI. 11). The digestive system is characterized by a very long oesophagus, that opens directly behind the proboscis-pore. The mouth is, however, quite separate from the rhynchodaeum. In the brain region the narrow oesophagus is situated between the brain-lobes (fig. 1, PI. 11), though its continuation into the gastric cavity takes place behind the cerebral organs (fig. 3, PI. 11). The gastric cavity with its richly folded epithelium and wide cavity is rather short (fig. 4, PI. 11), about twice the

6 172 GERARDA STIASNY-WIJNHOFF length of the brain, the enteric blind-gut accompanies it in about half its length; a very short pyloric tube opens into the intestine proper. Keferstein took the beginning of the gastric cavity to be the mouth (Keferstein, PI. v, fig. 10). The gastric cavity is accompanied by longitudinal muscle-fibres, that only are found at the inner side of the inner circular musculature (fig. 4, PI. 11). In the pregastric nephridial region, where the digestive system is represented by a rather narrow gastric tube a number of longitudinal muscles are found in the central body-parenchyma (fig. 3, l.m., PI. 11) of which one pair is particularly conspicuous (fig. 3, g.m., PI. 11). These muscles lie at the side of the rhynchocoelomic cavity, under the place of origin of the diverticula; they do not leave this position and go straight to the dorsal horns of the gastric cavity proper, where they lose their independence and become part of the longitudinal musculature of the cavity (fig. 4, PI. 11). I call them musculi gastrici. They are present in other species of this genus, though not always as conspicuous as here. The intestine proper has paired diverticula, that do not branch (Keferstein, PI. v, fig. 15) and are separated from the body margins by the rhynchocoelomic diverticula. The blood-vascular system shows the usual structure (Keferstein, PI. v, fig. 10). The rhynchocoelomic vessel enters into the cavity at the end of the ventral brain-lobes and leaves it already at the beginning of the gastric cavity, far before the enteric blind-gut (fig. 4, PI. 11). The nietamerie anastomoses go all round the intestinal pouches. The nephridium is found at the side of the brain and the gastric cavity. The excretory canal turns forward and opens near the beginning of the gastric cavity (fig. 3, PI. 11). The canal is found at the outer side of the nerve-cords and has a ventral pore. The g o n a d s are V-shaped and do not reach into the gastric region. The gonoducts open on the ventral side, just outside the nerve-cords, in both males and females. Habitat. Punnettia splendida is known from both sides of the Channel; it has been found on fresh oysters at St.

7 NORTH ATLANTIC POLYSTYLIFERA 17S Vaest la Hough; near Guernsey it lives in fathoms, near Roscoff at m. depth. Riches and the author found it at Plymouth. Piinnettia splenclida is characterized by the position of the mouth, the long oesophagus and the short pylorus, the number of proboscidian nerves (24-25), the well developed musculi gastrici, the situation of the rhynchocoelomic vessel, the position of the cerebral organs and the cerebral pore, &c. It differs from the Mediterranean representative of this genus in all these characters, as Punnettia hubrechti has its mouth beneath the brain, possesses a wide and short oesophagus, a very long pylorus, no enteric blind-gut at the side of the gastric cavity, twenty-two proboscidian nerves; the cerebral organs lie behind the brain, the canal opens into the middle longitudinal cephalic slit, the rhynchocoelomic vessel reaches from the end of the dorsal ganglia to the beginning of the pylorus, the first pair of rhynchocoelomic diverticula are found behind the nephridial mass, &c, &c. There can be little doubt that the Atlantic and the Mediterranean representatives constitute two different species of this genus. UNIPORUS BOREAMS Punnett. Figs. 6-11, PI. 12. Punnett. 'Proc. Zool. Soc, London', 1901, ii, p. 95. Drepanophorus borealis. 'Bergens Museums Aarbog', 1903, no. 2, p. 32. Drepanophorus borealis pr. p. Brinkmann. Ibid., 1915, no. 6, p. 13. Uniporus borealis. The following description has been made on the original sections of Punnett. The specimen, as mentioned by Punnett (1901), was obtained from Davis Strait. The external features are found in his description (1901, p. 95). Though the presence of head-furrows is confirmed by the structure of the epithelium (fig. 8, PI. 12), I have not been able to find the secondary grooves, which in some Drepanophori are so obvious. They seem here to be absent. As to the presence of eyes I have not found any, though

8 174 GERARD A STIASNY-WIJNHOFF Punnett states that four eyes are present. As the other species of this genus have no eyes, I believe his statement to be erroneous. Body-wall. The integument has been preserved very unsatisfactorily, but there can be no question as to the presence of a small number of gland-cells. The basement membrane is very thick and shows the same two layers, described in Drepanophorus ritteri Coe, only the cup-like lamellae stain less deeply after Punnett's method of staining; a difference of colour is, however, present as the limiting membrane is blue and the other parts are reddish. The presence of the head-furrow is indicated by a local thinning of the membrane. The circular muscle-layer is less thick than the basement membrane and shows very clearly an inner diagonal layer. The inner longitudinal muscle-coat has been developed more strongly, as all reproduced sections show. In the head-region (figs. 7 and 8, PI. 12) it is rather thin, and then increases enormously in the brain-region by the fibres of the precerebral septum (fig. 8, PI. 12). The thickness of this layer diminishes in the enteric region. In the lateral margins it is thinner than on the dorsal and ventral sides. The dor so-ventral musculature is well developed in the enteric region, as well between the enteric pouches as at the margins of the body (fig. 11, PI. 12). The body parenchyma is abundantly present in all parts of the body, and shows a number of irregular cavities, especially in the neighbourhood of the longitudinal musculature (figs. 8-10, PI. 12). The proboscidian system opens into an atrium, together with the digestive part. A well-developed rhynchodaeum leads to the insertion of the proboscis (figs. 7-8, PI. 12). The wall of the atrium has an extremely thin muscular coat, consisting of the same layers as the body-wall, with a high epithelium but without basement-membrane. Immediately behind the beginning of the oesophagus a thick circular muscle-layer indicates the presence of the rhynchodaeum (fig. 7, PI. 12). This muscle-layer is the forward continuation of the sphincter. The insertion of the proboscis takes place before the brain (fig. 8, PI. 12). Punnett says (p. 96): 'The proboscis is well developed and of about the same length as the body. Its epithelium is

9 NORTH ATLANTIC POLYSTYLIFERA 175 raised up into large papillae and through the extremely thick basement membrane upon which they rest, may be traced strong nerves entering their bases. The proboscis contains fourteen nerves and is attached near the end of the body to the ventral wall of the proboscis-sheath.' I can add: the new circular muscle-layer fails; only the outer circular layer and the longitudinal layer are present (fig. 6, PI. 12). The proboscidian nerves are given off from the ventral brain-lobes, two, a dorsal and a ventral one on each side. They form a ring just outside the thin musculature of the rhynchodaeal sac and from this ring the fourteen nerves originate and are directed to the insertion of the proboscis, where they enter each separately (fig. 8, PI. 12). The nerves in the first part of the sheath are found underneath the endothelium. The proboscidian sheath is connected with the muscular wall of the body by a very thick septum before and in the region of the brain (fig. 8, PI. 12), in the same way as has been figured by Brinkmann in Uniporus acuto caudatus (PI. ii, fig. 16a., and on the ventral side at the level of vie). In the wider part of the rhynchodaeum the sphincter is extremely thin, and has disappeared absolutely at the insertion of the proboscis. The outer part of the longitudinal fibres continue on the outside of the sphincter, but the bulk of them enters the proboscis, as the sections show very clearly. Behind the longitudinal musculature the nerves enter, and then the rhynchocoelomic endothelium appears resting on a thin layer of connective tissue which surrounds the proboscidian nerves. The longitudinal fibres of the septum are found outside them. This part of the proboscidian cavity is devoid of a muscle-coat. The first muscle-fibres appear inside the proboscidian nerve-ring and are circular. The first longitudinal fibres appear between them in the brain-region, behind the dorsal commissure; they lie mostly on the ventral side and are proximally rather scarce, on the sides they dwindle even more distally. The muscular coat of the rhynchocoelomic cavity is thin (Punnett, Textfig. 2). _ The diverticula are richly branched and do not communicate with each other. The first pair of diverticula originates in the region of the paired pouches of the blind-gut, though the

10 176 GBBARDA STIASNY-WIJNHOFF branches reach forward to the nerve commissures and even before and on the ganglia and cerebral organs (figs. 8, 9, 10, PI. 12). The presence of muscular fibres is very conspicuous in the contracted parts; in the extended branches, however, they are often difficult to recognize, especially if the picric acid has disappeared. The precerebral region shows a rather well developed muscular wall and a thin basement membrane; the proximal part is devoid of muscle-fibres in the centre (fig. 7, PI. 12), which consists of connective tissue with a great number of irregular spaces. The glands, described by Brinkmann for the other species of the genus, are also present, and lie at the dorsal side. In the margins a few longitudinal fibres are found coming from the precerebral septum and ending in the body-wall. The longitudinalfibreson the dorsal side of the rhynchodaeum continue into the body-wall along the atrial mouth and originate also together with the septum. The oesophagus is very wide and in contact with the rhynchodaeum. I have found no trace of eyes or other sense organs in this region. The nervous system shows a pair of cerebral ganglia that have a very unusual feature for the Reptantia. They (fig. 9, PI. 12) are situated in a ventral plane, underneath the rhynchocoelomic cavity and are connected by long commissures. The ventral commissure has no lining of ganglionic cells, and is thick and straight though the oesophagus is in close contact with its ventral side. The dorsal commissure is long and lies behind the ventral one; as the proboscidian nerves form a commissure directly after their origin, the dorsal brain commissure is the second one; it is thicker and originates from the dorsal ganglia whereas the proboscidial commissure originates from the ventral lobes. Because of the contraction of the rhynchocoelomic cavity the dorsal commissure has a number of folds and is not much curved. The dorsal ganglia are situated outside the ventral and are not longer than these; outwardly they are not well separated and even in section the dorsal ganglia are not much larger than the ventral ones. The fibrous core shows no differentiation of the two parts, as known in Paradrepanophorus, and the

11 NORTH ATLANTIC POLYSTYLIFBRA 177 bifurcation of the dorsal lobe is also absent. When the dorsal and ventral ganglia separate (fig. 9, PI. 12), the dorsal becomes the cerebral nerve, which immediately enters the cerebral organs and the ventral continues after a few sections into the lateral nerve. The inner neurilemma enters rather late, behind the point of disappearance of the cerebral nerve. As dorsal and ventral ganglia proximally reach to the same plane (fig. 8, PI. 12) the ventral lobes in reality are longer than the dorsal ones. Neuroehords are present. The lateral nerves are the direct continuations of the ventral ganglia and keep almost the same place as these throughout the body (figs. 10, 11, PI. 12; Punnett, Text-fig. 3). The distance which separates them is only one-third of the breadth of the animal. The cerebral sense organs (figs. 9, 10, PL 12) open to the exterior by a ventral pore at the level of the dorsal brain commissure (Punnett, Text-fig. 2). They are longer than the brain, and have a very irregular shape as the glands which open into the cerebral canal have only partly been absorbed by the organ. The pore opens immediately into the large organ, which is situated near the margin of the head and only touches the brain at the spot where the cerebral nerve enters into it (fig. 9, PI. 12). The rumen of the porus leads to the sac, which has folded walls and is rather narrow. This sac forms the proximal and dorso-lateral part of the sense organ; on its central side the nervous mass is found. The cerebral canal originates from the ventral wall of the sac and is directed backwards and inside; the ganglionic part is found between the sac and the canal. Several bundles of gland-cells are found outside and distally to these parts (figs. 9, 10, PI. 12); at the end they lie on the inner and dorsal side, where they constitute a part of the cerebral organ (fig. 10, PI. 12). Long slender masses of glands, which are the continuation of these parts are found laterally, one also ventrally (fig. 9, PL 12). After the disappearance of the sac and the absorption of the cerebral nerve the cerebral canal makes its crescent-shaped curve into the glandular mass and comes to lie in the centre, with two large masses of glands on the dorso- and ventro-lateral sides. The free bundles of glands maybe separated from the organ by dorso-ventral muscle-strands.

12 178 GEBARDA STIASNY-WIJNHOFF A large pigment ball is found between the glands, and some pigment may also be present in the free bundles. The peripheral situation and the absence of a free cerebral canal, the presence of several free glandular masses and the irregular shape of the organ seem to show a primitive state of development. Cephalic furrows are indicated by the thinning of the basement membrane (fig. 8, PI. 12), but secondary grooves are probably absent; it is not certain whether the cerebral pore opens into the furrows or not. The digestive system opens into the atrium through an unusually wide oesophagus (figs. 7, 8, PL 12) which is in contact with the rhynchodaeum and lies in the body-parenchyma on the musculature of the ventral body-wall. Underneath and immediately behind the ventral brain commissure it widens out to the gastric cavity, which, however, is very small and shows no diverticula; it is a narrow cavity with a few folds of the glandular epithelium (fig. 9, PI. 12). It is more or less triangular in section and not much broader than the rhynchocoelomic cavity, whereas the oesophagus is crescent-shaped in section. The first traces of the unpaired blind-gut reduce the dorso-ventral diameter of the gastric cavity, and with it appears the pylorus which is longer than oesophagus and gastric cavity together. A pair of large diverticula originate from the projdmal end of the blind-gut, which reach to the end of the cerebral sense organs (fig. 10, PI. 12), but not to the brain. The other diverticula show the same peculiarities as have been described by Brinkmann for Uniporus hyalinus(fig. 11, PI. 12). Each diverticulum shows three parts, a small ventral branch (v.ent.div.), a larger dorsal one (d.ent.div.) and a wide lateral sac; the ventral branch is separated from the others. The lateral sac is much more branched than in hyalinus, probably in consequence of the broader margins of the body, but the special lobe we described in Drepanogigas albolineatus is not present in this species. The vascular system shows a peculiarity in the region of the head which may be due to contraction; the cephalic anastomosis (fig. 7, PI. 12) takes place on the proximal wall of the

13 NORTH ATLANTIC POLYSTYLIFERA 179 atrium, and the blood-vessels are directed forward. At the tip of the snout they curve outward and then follow the margins of the head till they at once curve inwards to enter the brainring between the dorsal and ventral lobes (fig. 8, PI. 12). A short blind vessel to the cerebral sense organ conies off from the cephalic loop. The cerebral anastomosis is behind the ventral brain commissure ; the lateral vessels keep always the same place between the dorsal and ventral ganglia; the dorsal blood-vessel enters the rhynchocoelomic cavity at the end of the brain and leaves it in the course of the pylorus, which means that it is very long. The characteristic displacement of the lateral vessels in the nephridial region is absent (fig. 10, PI. 12). There are no metameric anastomoses. The nephridia are voluminous and very large. The nephridiopores are ventral (fig. 10, PI. 12), at the outside of the side nerves at the same level as the end of the cerebral organs and the beginning of the diverticula of the blind-gut. They lie in the first part of the nephridia, for the first sections are found behind the cerebral nerves and brain, and the nephridium reaches far into the pyloric region, even behind the end of the rhynchocoelomic vessel. The g o n a d s have been described extensively by Punnett; they ' are large and arranged in three rows on either side, which all open on the dorsal surface. They take the form of elongated tapering sacs, the widest portion of which is near the ducts' (1901, p. 96). The ovaries contain many eggs, the development of which has been described and figurated by Punnett. That Drepanophorus borealis Punnett belongs to Brinkmann's genus Uniporus seems to be quite certain. It differs from his two species in several respects, as already has been noted by Brinkmann; very instructive are the transverse sections of the three species, given by Brinkmann. HUBRECHTONEMERTES LANKESTERI Hubrecht. Figs , PL 18. Hubrecht (1886). "Report on the Nemertea", '"Challenger" Rep. Zool.', vol. 19, p. 18. Drepanophorus lankesteri.

14 180 GBRABDA STIASNY-WIJNHOFF Brinkmann (1915). 'Bergens Mus. Aarbog', no. 6, p. 26. Drepanophorus lankesteri. (1917). "Die pelagischen Nemertinen", 'Bergens Mus. Skr. N.R. 1, Bd. iii, p. 146 a. foil. Drepanophorus lankesteri. Stiasny-Wijnhofi (1923). 'Quart. Journ. Micr. Sci.', vol. 67, p Drepanophorus lankesteri. This specimen was caught by the ' Challenger' Expedition at Station 49 near the American coast. It was dredged in the waters of Nova Scotia, and probably lived at a depth of 85 fathoms between stones and gravel (43 3' 0" N., 63 39' 0" W.). Hubrecht's description of the external features shows no distinguishing characters. He writes (p. 18): ' As to its colour when alive the spirit specimen allows no other certain conclusion than that the dorsal surface is darker than the ventral, which may have been whitish. No special markings are now traceable on the dorsal integument and we may thus surmise that its natural colour which has been only partly preserved in spirit, was in life brown or red.' The one specimen measured 80 mm. in length and 3-5 mm. in breadth. The sketch (PL i, fig. 22) reveals no peculiarities at all. Hubrecht's collection of Nemertines contains the sections of this animal. They are series of longitudinal sections through the proximal part (containing the mouth and cerebral region as well as the greater part of the stomodaeal region); 2 a slide with transverse sections through the pyloric and enteric region, and 3" a slide with transverse sections through the tail. As several anatomical details described by Hubrecht have raised discussion (Brinkmann, 1915, 1917) a more detailed study of the original sections seemed of interest. Body-wall. The structure of the body epithelium has been elucidated very well by Hubrecht's PL x, fig. 2, and needs no comment. It is due to the staining with carmine that the cuplike basement membrane of the epithelial gland cells does not show up so clearly as f. inst. in Coe's D. ritteri, for it is present here just as well as in most other Polystylifera Beptantia. The basement membrane is rather thick (figs. 12, 17, PL 13) and is sharply separated from the thin outer circular muscle-layer;

15 NOBTH ATLANTIC POLYSTYLIFBRA 181 a single layer of diagonal muscle-fibres is present. The inner longitudinal musculature is not very well developed, and is extremely thin in the lateral margins (fig. 17, PI. 13; Hubrecht, PL ix, figs. 1 and 2; PL x, fig. 2). Everywhere in the body rather a large amount of parenchyma is present which envelops all organs and separates them from the body-wall. The snout is rather contracted, and seems to contain not much parenchyma (fig. 12, PL 13). The muscular wall is thin; the central part, however, is nearly filled up by the large eyes, the head-gland, the atrium with oesophagus and rhynchodaeum, and the longitudinal muscles, nerves, and blood-vessels embedded in the parenchyma. The presence of an atrium (fig. 12, PL 13) into which both rhynchodaeum and oesophagus open has been noted before by the author (1923). The epithelium of the atrium seems to be the same as that of the snout, though it is lower; other differences were not observed. The wall is devoid of the thick basement membrane, and has no musculature of its own. The thick ventral longitudinal muscles, the musculi retr. capitis, are inserted in this place and accompany the oesophagus (fig. 12, PL 13). They are directly continued into the precerebral septum, that is pierced through only by oesophagus, rhynchodaeum, nerves, and cephalic vessels. The large eyes are arranged in lateral rows. I counted ±22 at each side, arranged in a shorter dorsal (with ±7 smaller eyes) and a longer ventral row with the largest eyes (±15) that are about twice the size of the smaller ones. Probably the dorsal rom' is constituted by a short proximal row of two to three eyes and a more distal cluster of four to five eyes. As the sections of the contracted snout are not exactly horizontal, it was impossible to ascertain this. The head-gland is rather conspicuous but short. Part of the gland has been figured by Hubrecht in PL xv, fig. 13. The rhynchodaeum has a thin basement membrane, but is devoid of a muscular coat. The sphincter is restricted to its distal part and lies just before the precerebral septum. The insertion of the proboscis takes place immediately before the brain. The proboscis is absent.

16 182 GERARDA STIASNY-WIJNHOFF The rhynchocoelomic cavity has the usual basket-like structure of the wall. It ends just above the anus. The first diverticula are found at the side of the end of the dorsal ganglia (fig. 15, PI. 13). All diverticula are simple unbranched tubes. The thickness of the muscular wall is due to contraction. Hubrecht's figure (PI. x, fig. 4) gives a false impression; whenever the wall is thin, the lumen of the diverticulum is dilated (fig. 15, PI. 13). The diverticula are inserted rather dorsally in the wall of the proboscis sheath; they take their way through the parenchyma to the dorsal wall of the lateral intestinal pouch and follow it to the ventral side. Here they bend distally (3 d rh.div., fig. 15, PI. 13), and may even lie at the side of or under the next intestinal pouch; in this way two neighbouring diverticula may come to lie side by side, but never do their lumina fuse (fig. 15, PL 13). The diverticula are very different in size. Some are short and end directly at the side or under the intestinal pouches to which they belong. Others, and this is usually the case with the next one, bend distally as described above and then return again to their original intestinal pouch. In the gastric region (fig. 15, PI. 13), where the intestinal pouches fail, the diverticula behave in the same way. The first diverticula f. inst. go straight to the margins, till they are behind the cerebral organs; here they bend distally and end as a thin-walled dilated pouch. The second pair, however, has a parallel course at first, and gets its distal bend a little more laterally. Then they return and lie at the lateral side of the first pair, and even reach the hind end of the brain. So they get more forward than the first pair of diverticula. The third pair keeps a more distal position. All this shows that Hubrecht made a mistake in supposing that some successive caeca intercommunicate by means of their short longitudinal extensions (p. 19). Brinkmann's supposition (1915, p. 26), that the 'dickwandige Khynchocoelom-divertikel jedenfalls etwas verastelt sein miissen', I am not able to support either. Neither transverse sections from the caudal or the enteric and pyloric regions, nor the horizontal sections from the more proximal part of the animal revealed any traces of branch-

17 NORTH ATLANTIC POLYSTYLIFERA 183 ing of the diverticula, as supposed by Brinkmann, though I sought for them purposely and made reconstructions. Each intestinal pouch has its own rhynchocoelomic diverticulum. Nervous System. The study of the structure of the brain of our only specimen was very much thwarted by the unusual way in which the parts were sectioned. They are cut not horizontally but more obliquely, and as we are used to study the nervous system in transverse sections, many difficulties arise in recognizing several important features. Hubrecht gives in PL ix, fig. 10, a reconstruction of the nervous system with the cerebral organs. The position of these organs quite behind the brain seems not to be in accordance with the facts. Many sections give this impression; but we must remember that each section shows only a certain part of the brain, and that the brain consist of several parts with different extensions. The ventral brain-lobes are much shorter than the dorsal ones, and we know that the dorsal brain-lobes have acquired in almost all Polystylifera Eeptantia a special distal continuation that ends in the cerebral nerve (fig. 13, PI. 13). This distal part is present in Drepanophorus lankesteri and is exactly as long as the cerebral organ, that lies between this lobe and the nerve-cord or the ventral ganglion. The more ventral position of the cerebral organ is the cause of Hubrecht's mistake. As, however, our sections are not quite horizontal, on one side the true position of these organs is revealed very clearly; on the other side of the animal the distal brain-lobe and the cerebral organ, however, can never be seen in one section. The position of the distal end of both cerebral organs and dorsal brain-lobes is fixed by the extension of the gastric cavity. This dorso-ventral wall of the gastric cavity demarcates the end of both organs, and shows very clearly that they lie side by side; the dorsal brain-lobes lie more dorsally and medially, the cerebral organs laterally and more ventrally, filling the space between lateral nerves and the dorsal part of the brain. Dorsal and ventral ganglia have as usual an outer neurilemma but no inner neurilemma. They give off a number of nerves that pass to the eyes, one pair of proboscidian nerves, one pair of oesophageal nerves, and probably one pair of subdorsal NO. 306 O

18 184 GERARDA STIASNY-WIJNHOFF nerves. The dorsal ganglia are longer than the ventral ones; they are not separated from each other. The dorsal commissure has longitudinal folds because of the contraction of the rhynchocoelomic cavity. The ventral commissure is much shorter and seems not to be exceedingly stout. It is a straight commissure. Neurochords are present ('Challenger' Eeport, PI. xiv, fig. 9). The dorsal brain-lobes show two parts of which the inner and more proximal one runs directly into the ventral lobe and receives the nerve-fibres from the dorsal commissure. The distal and more peripheral part gives off only one nerve, the cerebral nerve. This fibrous mass does not bifurcate. Between this part and the ventral ganglion extensive direct relations cannot exist. The ventral ganglion is small; the fibrous mass runs into the ventral commissure, distally, however, a bifurcation takes place; the inner branch runs into the side nerves; a continuation of dorsal fibres into the nerves seems to be absent. The position of the side nerves seems not to vary in the different regions: immediately after their origin they are found at the inner side and under the cerebral organs, surrounded on all sides by the body parenchyma. This place they keep in the stomodaeal region, and even in the caudal part they lie on the ventral side, but rather laterally. If ever they are in contact with any organs, these are the intestinal diverticula, but never do they lie near the muscular body-wall (fig. 17, PI. 13). The peripheral nerves are very conspicuous, they are, however, by no means so regularly situated as Hubrecht figures them in his reconstruction, and I have not been able to find commissures joining the ventral branches. In respect of this and also of the relative situation of cerebral organs and brain I believe Hubrecht's reconstruction of the nervous system to be incorrect (PI. ix, fig. 10). The sense organs are well developed; more than forty eyes are arranged in the usual way (cf. p. 15); cephalic furrows are present. The cerebral pore opens into a short semicircular epithelial furrow. Whether longitudinal furrows as in most Reptantia are present, I could not decide. The behaviour of basement membrane and epithelium seems to indicate it. The cerebral organs lie very superficially; as the longitudinal muscula-

19 NORTH ATLANTIC POLYSTYLIFERA 185 ture is extremely thin in this region, a cerebral canal is absent, and the organ itself opens directly to the exterior (fig. 13, PI. 13). The cerebral canal divides into two branches; the lateral sac is narrow and shows no folds in its epithelial lining; the medial branch is short and ends blindly in the ventro-median and distal glandular part. The fibrous nervous mass of the ganglion is found between both. The digestive system is independent from the rhynchodaeal system; they both open into a common epithelial atrium (fig. 12, PI. 13). The oesophagus is narrow, surrounded by the longitudinal musculature of the head. The gastric epithelium shows itself at first behind the cerebral ring, in the region of the cerebral organs. In this region the epithelium is thrown into numerous folds (fig. 15, PI. 13); the gastric cavity fills up the whole space between the hind parts of the ganglia and the side nerves, and cerebral organs. The gastric cavity is surrounded by circular muscle-fibres; directly behind the cerebral organs the gastric musculature is in contact with the muscular bodywall (muse, gastrici?). At this place the gastric cavity reaches its greatest transverse section, but then it diminishes in size and shows no folding of its epithelial wall. The pylorus, as I call this part, is still very wide at first and its dorsal wall is crescentshaped, surrounding the narrow rhynchocoelomic cavity even laterally (fig. 15, PI. 13). The blind-gut does not reach as far forward as the pyloric cavity, which is rather long. It is separated from the brain and cerebral organs by a distance of less than a brain length. At first the narrow blind-gut reaches somewhat nearer to the brain than its paired diverticula. In the space between these and the gastric cavity lies the nephridial organ. The first pair of diverticula is very short and narrow, and lies at the inner side of the second pair. The others have the same structure as the enteric diverticula. These are unbranched and wide and show no metameric arrangement. Though they are paired, in horizontal sections (fig. 16, PL 13) may be observed that two kinds of diverticula occur: a shorter more median and dorsal branch and a lateral longer sac that reaches over the lateral nerves (fig. 17, PL 13), and which shows a tendency to bifurcate (fig. 16, PL 13). They both open into the iptestine

20 186 GEBARDA STIASNY-WIJNHOFF proper and are separated by strong septa; Hubrecht's PI. ix, fig. 1, therefore, is a mistake, as they are not two branches of one diverticulum but two separate diverticula. As a rule they alternate, but the horizontal sections show quite clearly that exceptions are numerous. That it is not a case of more or less room is shown in the tail, where there is plenty of parenchyma and still the two kinds of diverticula are present. Each diverticulum is accompanied by a rhynchocoelomic diverticulum (fig. 16, PI. 13), but the vascular anastomosis is only present above the lateral ones. Bach 'metamere' therefore contains two rhynchocoelomic and two intestinal pairs of pouches on one vascular anastomosis; the gonads are found between the intestinal diverticula. The blood-vascular system has the usual structure. The cephalic loop shows no peculiarities; the anastomosis lies before the sphincter rhynchodaei. The cerebral anastomosis gives rise to a dorsal and two lateral vessels; the dorsal one enters almost immediately into the rhynchocoelomic wall (fig. 15, PI. 13) and leaves the cavity at the beginning of the blind-gut. The anastomoses are absent in the pyloric region, but start immediately behind the mouth of the pylorus into the enteron (fig. 17, PI. 13). They do not encircle the intestinal pouches but pass between them, behind each lateral branch and before each dorsal one; half as many anastomoses are present as pairs of diverticula or gonidial rows. The nephridium consists of wide channels and lies between the cerebral organs and the enteric diverticula (fig. 15, PI. 13); a few channels are present in the pyloric region too. The efferent duct passes outside the nerve-cord in the beginning of the blind-gut and takes a more or less winding rather long distal course (fig. 14, PI. 13). One pair of nephridiopores is found; they lie probably half-way down the blind-gut; but, as part of the sections are horizontal and the others are transverse, I can only guess the distance from the pyloric mouth. The gonads are very small and have not yet developed pores. They lie at both sides of the lateral nerves, are only ventral in position, and contain each one egg. Under each pair of intestinal diverticula are found two pairs of ovaria, a medio-ventral and

21 NORTH ATLANTIC POLYSTYLIFEEA 187 a latero-ventral one. This is very curious, as such a position and so few eggs are only known in Polystylifera pelagica. There is no question that Hubrechtonemertes belongs to the non-pelagic Polystilifera. The structure of the rhynchocoelomic cavities, of the brain, and cerebral organs, of the digestive tract, blood-vessels, and nephridium is much the same as in Paradrepanophorus. Though the atrium and the number of gonads show the same features as in the genus Uniporus, the presence of eyes, the structure of the cerebral organs and brain, the unbranched rhynchocoelomic diverticula and enteric pouches, the ventral position of the gonopores and the metameric anastomoses tell against a nearer relationship between the two genera. Moreover, we know other genera with an atrium, such as Drepanogigas albolineatus (Burger), Paradrepanophorus corallinicola Wijnh. and several indo-pacific species. Of these Drepanogigas shows other features in common with Hubrechtonemertes, for instance, the ventral position of the brain, the absence of a cerebral canal, the development of dorsal and ventral intestinal diverticula, though these are inserted somewhat differently, the position of the metamerical blood-vessels in regard to these pouches, the small ovaries with few eggs, that open separately in our species and into a common bag in Drepanogigas; but in both cases on the ventral side. The differences in structure of brain and cerebral organs, rhynchocoelomic diverticula, gonads, and digestive tract are, however, too great to admit of a nearer relationship. Close affinity to the West Indian genus Polys c h i s t a, with its peculiar cephalic vessels, branched rhynchocoelomic diverticula, strong inner circular muscle-layer, glandular processus of the cerebral organ, &c, seems out of the question. A comparison with Paradrepanophorus reveals, however, many differences too. These are: the shape and number of gonads, the position of the nephridiopore (proximal in Paradrepanophorus, distal in Hubrechtonemertes), the position of the anastomoses, the bifurcation of the fibrous mass of the dorsal ganglia in Paradrepanophorus, the position of the cerebral organs to the brain. Though Paradrepanophorus seems to be nearly related,

22 188 GERARDA STIASNY-WIJNHOFF it is, however, certain, that Drepanophorus lankesteri cannot be included in one of the existing genera of Drepanophoridae or Reptantia. I call it therefore Hubrechtonemertes in honour of my deceased teacher, who first described this species. EXPLANATION OF PLATES 11, 12, 13. LETTERING. bas.m., basement membrane; bl.g., blind-gut; ceph.bl.v., cephalic bloodvessel; ceph.n., cephalic nerve; cer.can., cerebral canal; cer.n., cerebral nerve; cer.org., cerebral organ; cer.p., cerebro-porus; cer.s., cerebral sac; circ.m., inner circular musculature; cm., circular muscle-layer; d.bl.v., dorsal blood-vessel; d.ent.div., dorsal enteric diverticulum; d.g., dorsal ganglion; dist., distal; dors.c, dorsal brain commissure; d.v.m., dorsoventral musculature; end., endothelium; ent., enteron; enl.div., enteric diverticulum; ep., epithelium; g.cav., gastric cavity; gl.can., glandular canal; g.l.m., gastric longitudinal musculature; gl-pr., glandular processus of cerebral organ; g.m., musculus gastricus; i.l.m., inner longitudinal musculature; l.bl.v., lateral blood-vessel; l.ent.div., lateral enteric diverticulum; l.m., longitudinal musculature; l.n., lateral nerve; m.relr.cap., muse, retractor capitis; nephr., nephridium; nephr.can., nephridial canal; nephr.por., nephridiopore; n.l, nervous layer; o.c.m., outer circular muscle-layer; oes., oesophagus; oes.n., oesophageal (gastric) nerve; ov., ovary; pap., papilla of proboscidian ep.; par.cav., parenchymal cavity; prec.s., precerebral septum; prob., proboscis; prob.n.c, proboscidian nerve commissure; prox., proximal; pyl., pylorus; pyl.cav., pyloric cavity; rh.c, rhynchoeoelomic cavity; rh.d., rhynchodaeum; rh.d.cav., rhynchodaeal cavity; rh.div., rhynchoeoelomic diverticulum; rh.ep., rhynchodaeal epithelium; rhynch.bl.v., rhynchoeoelomic blood-vessel; rhynch.oav., rhynchocoelomic cavity; rhynch.d., rhynchodaeum; rhynch.div., rhynchoeoelomic diverticulum; rhynch.w., rhynchoeoelomic wall; s.d.n., subdorsal nerve; sphr.rh., sphincter rhynchodaei; v.ent.div., ventral enteric diverticulum; v.g., ventral ganglion. PLATE 11. Punnettia splendida Keferstein. Kg. 1. Transverse section of brain-region, showing the cerebral pore and the dorsal commissure; dorsal and ventral ganglia are not separated. Fig. 2. Sagittal section through the brain and the cerebral organ with its distance to the basement membrane. Fig. 3. Transverse section through the nephridial region and nephridioporus, showing the musculi gastrici and some smaller gastric longitudinal muscle-bundles.

23 NORTH ATLANTIC POLYSTYLIFERA 189 Fig. 4. Transverse section through the gastric region with inner circular muscles. Kg. 5. Transverse section of the enteric region with enteric and rhynchocoelomic diverticula. PLATE 12. Uniporus borealis Punnett. Kg. 6. Part of a transverse section through the ejected proboscis. Fig. 7. Transverse section through the snout behind the atrium, showing the large blood-vessels and thin musculature. Fig. 8. Transverse section in the region of the precerebral septum, before the insertion of the proboscis, showing the commissure of the proboscidian nerve and the thin-walled rhynchodaeal sac behind the sphincter, a large number of cavities in the parenchyma and the way in which the blood-vessels enter into the brain-ring; rhynchocoelomic diverticula; the cephalic furrows are marked by a +. Fig. 9. Transverse section through the gastric and cerebral region, demonstrating the structure of these organs and of the cerebral organs with many bundles of gland-cells. Fig. 10. Transverse section through the region of the nephridiopore; enteric diverticula on the cerebral organs, large nephridia, pylorus, branched rhynchocoelomic diverticula. Fig. 11. Transverse section through the enteric region, position of the branched enteric diverticula, the ovaries, and the dorsal blood-vessel. PLATE 13. Hubrechtonemertes lankesteri Hubrecht. Fig. 12. Nearly horizontal section through the snout, showing the structure of the atrium, the origin of the oesophagus and the insertion of the mm. retractores capitis. Fig. 13. Horizontal section through the doraal-ganglion and the cerebral organ with its connexion with the body-wall; large lateral lobe of the dorsal ganglion with cerebral nerve. Fig. 14. Horizontal section through the nephridial canal with distal porus. Fig. 15. Left part of nearly horizontal section, showing the first rhynchocoe.iiocoelomic diverticula near their origin at the right side of the rhyncho- :oelomic cavity, the left horn of gastric cavity, and pylorus at the left dide, with the section of dorsal ganglion, nephridium, rhynchocoelomic diverticula, and enteric pouches outside of them in a more ventral plane. The short first enteric diverticulum does not reach to this level. Thinwalled end of the first two rhynchocoelomic pouches.

24 190 GERARDA STIASNY-WIJNHOFF Fig. 16. Part of a horizontal section in the pyloric region, showing the pouches of the blind-gut, the septa, and the rhynchocoelomic diverticula; beginning bifurcation of the lateral enteric pouches. Fig. 17. Transverse section of the pyloric region, demonstrating the insertion of the rhynchocoelomic and enteric diverticula and the position of lateral nerves and blood-vessels.

25 Quart. Journ. Micr. Sci. Vol. 77, N. 8., Pi 11 T p dors c rhynch w gl pr cer org bas m cer can + /.rhynch cav \ cer s I ceph n G. Stiamy-Wijnhoff, del.

26 Quart. Journ. Micr. Sci. Vol. 77, N. S., PL 12 /\rh d cav oes Sph rh v 9 nephr L bl v nephr e R J m i L m rhunch cav v enb div -rhynch bl > d v m LbLv G. Stiasny-Wijnhojf, del.