Cornell aniversity Library THE GIFT OF AG 376.2 3.0./61b4.. - 1287 RETURN TO ALBERT R. MANN LIBRARY ITHACA, N. Y. “ TINTERLIE:.- PAN DATE DUE 4 GAYLORD TT Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924003409327 STUDEES. ‘ON THE ECTOPARASITIC TREMATODES © OF TAPAN BY - SEITARO GOTO MHgakushi ‘Reprinted front the ie of the Coltexe of Bigience, Imperial —_ University, Tokyo 3 ‘forming Part Z # Vol. "RLT TOKYO. 1894, {The 27th Year, Meiji.] 39] ast Guy A.\4 abe Studies on the Ectopacasitic Trematodes of Japan, mn by ‘Seitaro Gotd, Rigakushi, \ no i Science C pllege, Imperial University, Tokyo. \ ; With Plates I-XXVII. ‘Introduction. It was originally intended that I should publish these studies jointly with Prof. Ijima; but as he is engaged on another work, it has become necessary for me to take the whole responsibility upon my- self. The species on which these studies were made were for the most part collected by myself from various parts” of Japan during the sum- mers of 1889, 90, 91, and "92. The present part does not include the Gyrodactylide, the study of which I am still prosecuting ; and as this will occupy me for some years longer, I have thought it advisable to publish what is ready now, particularly as I have already been able to make out the general anatomy of some of the Gyrodactylide, and can thereforé take them into account in judging of the natural affinities of the different species. The specimens collected by me were usually killed with hot saturated solution of corrosive sublimate. This reagent gives, so far as my experience goes, the best genera] result, fixing the worms usually in an outstretched condition and thus facilitating the process of sec- : 1). The collection was made at the following localities: Hakodaté, Misaki, Tolkya, Mitsugahama (in Tyo), Ujina (the port of Hiroshima), Hagi, and Mogi (near Nagasaki). 9 8. GOTO. = tioning. Moreover, corrosive sublimate can be sv easily carried about, and its saturated solution so easily made that it'is, generally speaking, by far the best reagent for use on a collecting tour. The specimens were preserved in 70 °/, alcohol. For staining sections I have almost exclibively used Kleinen- berg’s solution of haematoxylin. I have tried picro-carmin and borax-carmin, but they did not give good results, although the latter was very excellent for staining specimens mbunted in toto. I have also tried cochineal tincture so highly recommended by Lang for the glandular cells of polyclads; but it gave no differential staining whatever. For preparation in toto, the specimens were killed under the pressure of a cover-slip over the flame of an alcohol lamp, and were directly immersed in 70°/, alcohol, in which they were preserved together with other specimens. For staining I have used borax- carmin ; the over-stain being thoroughly washed cut with acidulated 70 °/) alcohol. In most specimens, only the internal organs and the nuclei of the mesenchyma remain stained, while the mesenchyma itself is wholly decolourised, so that the result forms altogether a very beautiful object under the microscope. To Prof. Ijima and Prof. Mitsukuri are due my warmest thanks both for supervision and for giving me suggestions and the most friendly assistance. To Profs. Parona and Perugia of Genoa, Prof. Monticelli of Naples, and Prof. Ramsay Wright of Toronto, I am indebted for their courtesy in sending me their papers on ectoparasitic Trematodes. Finally but not least my best thanks are due to the authorities of the Imperial University for taking charge of the publication of the paper. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 3 A. Anatomy and Histology. 1. External Form of the Body. Broadly speaking, the form of the body is that of the blade of a leaf with a rounded apex; and as the leaf varies from an orbicular to a linear or lanceolate shape, so the body of ectoparasitic Trematodes varies in form between the same extremes. In Microcotyle the body is generally elongated and lanceolate or fusiform, the posterior end being rather pointed. In some species of this genus the halves of the body are asymmetrical, one being longer than the other, so that the axis of the body forms a curve or even a erooked line; eg. in M. reticulata (PI. I, fig. 5) and M. seiaenae (Pi. TL, fie. 6), In cross-sections the body of Microcotyle presents the form of an ellipse, of which the minor axis becomes greater and the major axis much shorter as the section approaches the anterior end. In the pos- terior portion of the body where the suckers are present, the cross-section is often semicircular in outline, the diameter being the ventral side. In Axine (Pl. VID), one side of the body is always longer than the other, and the posterior portion of the longer side makes an angle with the anterior part, so that this portion ‘looks like the posterior margin of the body, and has actually been so regarded by preced- ing writers. But that itis really a part of the lateral margin of the body seems to me beyond doubt both from the presence of suckers on the other side and from the course of the principal nerves and the excretory vessels to be described further on. In A. abervans the sucker-bearing portion of the longer side is perfectly straight and makes an acute angle with the anterior part, so that one is tempted to. regard it as the posterior border of the body ; but here too there is a sucker on the other side. In accordance with the general asymmetry 4 8. GOTO. of the body, the posterior end has come to be situated quite laterally, and the general form of the body to be more or less triangular. An interesting fact relating to the asymmetry.of the body in Azine is that the longer side may be either the left or the right. Thus, in A. heterocerca, out of the nine specimens which I have examined for the purpose, three had the left side of the body shorter while in the remaining six the right side was shorter. This fact, though apparently insignificant, will be found to be of use at least as a check in judging of the value of some diagnostic characters given by previous workers. The cross-section of the body is in Aine band-shaped ; and the thickness of the body diminishes, while its breadth increases, as we proceed towards the posterior part, so that in this region the cross- section presents the shape of a narrow ribbon. In Diclidophora (Pl. X), the cross-section presents no great devia- tion in outline from that of Microcotyle, but the general form of the body is greatly modified by the fact that the four pairs of suckers are hemi- spherical, and are borne on the posterior margin of the body arranged in a semicircle or in a horse-shoe shape. Moreover, in many species each sucker is borne on a long pedicel (Pl. X, fig. 9), a feature evidently which has suggested the generic name of Octodactylus to Sir John Dalyell. The portion which bears the suckers, the “ Haftscheibe” of German authors and which I shall call the “caudal disc,” is in all species more or less distinctly: separated from the anterior portion by a constriction of the body. In Diclid. tetrodonis (Pl. X, figs. 1 & 2), however, the posterior portion of the body is con- siderably elongated, so that the transition to the caudal disc is more gradual. In Hesxacotyle (P]. XIII), the body is again much flattened, but its absolute thickness is very much greater than in the other genera. STUDIES OX THE ECTOPARASITIC TREMATODES OF JAPAN. 5 It is sharply pointed at the anterior end, very broad in the middle portion, presents a constriction at a short distance from the posterior end, again broadens out, and then suddenly diminishes in breadth, so that at this part the lateral borders form the two equal sides of a very flat isosceles triangle with its apex directed posteriad, and have generally been designated as the posterior border. ‘There is however a small notch at the apex. In Octocotyle (PI. IX) and Onchocotyle (Pl. XV), the body is com- paratively much thicker, and the cross-section presents in some parts almost a circle. In Octocotyle, the caudal disc is not distinctly dis- tinguishable from the rest of the body, the suckers being borne simply on the ventro-lateral margins of the posterior portion (PI. IX, fig. 7). In fig. 1 on Pl. IX, the caudal disc is ‘apparently set off from the remaining portion of the body by a sudden diminution of breadth ; but this has been caused by the specimen having been killed under the pressure of a cover-slip, and the body proper having been abnormally flattened in consequence. In fresh specimens or in those killed free, there is no such distinct boundary. In Onchocotyle, on the other hand, the caudal dise is distinctly marked off from the anterior part by a constriction, and bears, as is well known, a subcylindrical appendage projecting from its anterior end on the dorsal side of the body. The end of this appendage, which bears a pair of suckers (mistaken by Taschenberg for the terminal vesicles of the excretory system, see p- 28) and a pair of hooks, is in my opinion to be regarded as the posterior end of the body, with the body bent a little obliquely on itself towards the dorsal side, so that the suckers have come to lie apparently on the dorsal side. In proof of this view, it may be mentioned (1) that the two surfaces of the appendage and the caudal disc are seen, in serial sections, to be directly continuous with the dorsal and ventral ‘surfaces of the body ; (2) that there is a pair of hooks at the end of 6 8. GOTO. the appendage, which is usually the case in other genera; and (3) that according to this view the apparently dorsal side on which the suckers are borne, is really the ventral side, a fact in harmony with all the cases hitherto known. In Monocotyle (Pl. XVII) again, the body is much flattened, and its cross-section presents almost the form of a crescent whose inner side is ventral. In general outline the body is elongated, broad, and a little bordering on the oval. The posterior end is quite sharply pointed. -Anteriorly the body becomes narrower, but again somewhat broadens out in front, where at the end there is a large, shallow notch (PI. XVII, figs. 1 & 2). In Calicotyle (Pl. XIX), the shape of the body is that of an ovate, heart-shaped leaf, its apex forming the anterior end, and its basal notch bearing the posterior sucker (Pl. XIX, figs. I, 2, & 3). In Tristomum (Pls. XX—XXV), the body is mostly notched at the posterior end, and its form varies from that of an orbicular to that of an oval or ovate, heart-shaped leaf with apex more or less truncate ; the truncated border being sometimes convex, sometimes concave, and sometimes almost straight. In Epibdella, the body presents the same general outline as in Tristomum, except that the posterior end, instead of being notched, becomes gradually narrower and is directly continued into the sucker. 2. The Investing Membrane. In my paper on Diplozoon” I called this membrane the epidermis, assuming in so doing that it corresponds to the true epidermis of other animals. I have still no cause for recantation ; but as an antagonistic 1). This Journal, vol. IV, pt. 1, 1890. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. vi opinion has been advanced by Brandes,” and as the genesis of the membrane in question hag not been made out embryologically, I have thought it better to use a non-committal term, and have adopted for this purpose the name used by Wright and Macallum,” although it is in some respects not a very convenient one. Three layers can be distinguished in the investing membrane of the ectoparasitic Trematodes. These I shall call the cuticle, the sub- cuticle, and the basement membrane. These terms, I am well aware, are all preoccupied, and bear different significations according to different writers ; but the coinage of new words is not very desirable and is moreover not an easy task for one writing ina foreign language. Words hitherto in use can, however, be used in a new sense without any danger of occasioning confusion, when clear definitions are given. The term ‘ subcuticle” might be somewhat objectionable, as liable to be confounded with the “‘Subcuticularschicht” of Taschenberg and some other writers on Trematodes; but this is now so generally recognised to be nothing more than the cortical portion of the mesenchyma that there is, I believe, no serious danger of in- troducing confusion of ideas by adopting, in this paper at least, the terminology here proposed,” in place of ‘epidermal layer’ which I used in my former paper. The cuticle is a very thin, structureless, refractive layer which is very distinct in fresh specimens. In sections its existence is indicated 1). Brandes—Zum feineren Bau der Trematoden. Zeitschr. f. wiss. Zool., Bd. 53, 1892. p. 558. 2). Wright and Macallum—Sphyranura Osleri. Journal of Morphology, vol. I, 1887. p. 1. 8). Monticelli in a paper which was received after the above had been written, calls the investing membrane “ ectoderma,” and claims to have demonstrated in it the remnants of the original nuclei in the form of vesicles containing deeply stained corpuscles. Granting that the nuclei of the original epidermis may in some species remain in a comparatively unaltered state, it seems to me that the vesicles figured by Monticelli are too numerous to be regarded as the remnants of the nuclei of the ectoderm of the Cercaria, which are, according to the statements 8 8. GOTO. by a fine line on the external surface of the investing membrane, which usually stains in haematoxylin deeper than the subjacent layer ; but its thickness is always very insignificant, and can not be measured with any approximation to accuracy even under the magnifying power of 300 diameters. But that it is a distinct layer of cuticular nature is clearly proved by the fact mentioned in my former paper, that when any fresh specimen is observed in water under the cover- glass for a sufficient length of time, watery blisters are formed in various parts of the investing membrane, and the cuticle is raised from the subjacent layer. ’ Next to the cuticle comes a layer of varying thickness, form- ing my subcuticle. In its behaviour towards staining fluids, it is somewhat different in different species, the difference, however, lying only in the different intensity of its affinity with stains. For in- stance, in Microcotyle and Onchocotyle it is but slightly stained, while in Tristomum and Monocotyle it takes up the stain with greater avidity. In most species this layer is more or less granular, the ground- substance being formed by a uniformly stained, structureless substance. This ground substance seems, in the fresh state, to be of a semifluid. nature in most cases, and of a greater density than water. This I infer from the fact that when watery blisters are formed under the circum- stances already referred to, the water seems to pass into the subcuticle by a simple osmotic process and mix freely with its substance,—the substance of the subcuticle being quite undistinguishable from the water taken in. In Onchocotyle, the subcuticle appears striated in cross-section, the striation being caused by numerous fibrillar structures traversing it at right angles to its thickness. Besides these fibrils, of Schwarze, Schauinsland, and Biehringer, very few in number (Cf. Monticelli—Studii sui Trematodi endoparassiti: Primo contributo di osservazioni sui Distomidi. Spengel’s Zoolog. Jahrbiicher, III. Supple., 1893.). STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 9 granules are also observable. In all the other species, the subcuticle shows only granules, which are, however, more numerous in some than in other species. For instance, in Microcotyle, Axine, Monocotyle, Diclidophora, and Hezacotyle the granules are comparatively sparse, while in T'ristomum they are very numerous and exceedingly fine in certain regions, ¢.g. on the ventral side of the anterior suckers. Inward to the subcuticle comes the basement membrane. In cross- sections it is very deeply stained, and very distinctly separable from the subcuticle but somewhat less so from the subjacent mesenchyma. Its thickness can not be measured with any accuracy, but it is always thicker than the cuticle. Hallez? has shown with regard to plana- rians, that the basement membrane is of the same nature as the dense layer of connective tissue that surrounds the internal organs. Gene- tically therefore it belongs to the mesenchyma rather than to the investing membrane, and it is only in accordance with custom that I have described it as forming one of its layers. In Tristomum sinuatum and Trist. ovale there are numerous papillae on the surface of the body. These are of two kinds. Those of one kind are mostly perceivable with the naked eye. In Trist. sinuatum: they are confined to the dorsal surface of the body, and measure on the average about 0.028 mm. in height, the extremes being 0.014 mm. and 0.041 mm. ; the larger ones appearing to the naked eye as granulations. In Trist. ovale, on the other hand, they are confined to the ventral side where they are very numerous, and are much larger than in Trist. sinuatum, measuring on the average about 0.08 mm. in height, the extremes being 0.032 mm. and 0.122 mm. ‘The papillae of the other kind are all microscopic and are far less numerous than those of the first kind. In Trist. ovale they are 1). Hallez—Embryogénie des Dendrocoeles d’eau douce, 1887. p. 78. 10 S. GOTO. almost uniformly distributed on the dorsal surface, while in Trist. sinuaium they are mostly confined to near the lateral margins of the ventral side ; and in both species they measure on the average 0.011 mm. in height. ‘They are probably tactile organs. In sections the papillae of the first kind (PI. XXIV, fig. 2) are seen to be simple elevations of the investing membrane together with the underlying mesenchyma which presents, however, a somewhat different appearance from that of the more internal parts, and will be described under the mesenchyma. These papillae are always traversed lengthwise by- the terminal ramifications of the dorso-ventral muscular fibres, one of which usually ends at the very apex. The papillae of the second kind are distinguished not only by their minute size but also by the total absence of muscular fibrils ; and although I have not been able to demonstrate in them any nervous fibril I believe it will be found out by the application of appropriate methods. The total thickness of the investing membrane is in Microcotyle usually a little less than 0.005 mm.; in Axine heterocerca a little less than 0.004 mm. ; in Diclidophora 0.005 mm. ; in Hezacotyle acuta and Octocotyle minor 0.004 mm.; in Monocotyle 0.008 mm.; and in Onchocotyle 0.003 mm.—0.004 mm. It should however be borne in mind that the thickness of the investing membrane varies considerably in different parts of the body. Thus, in Tr. ovale it is 0.003 mm. on the ventral side and 0.012 mm. in the anterior part of the dorsal side ; again in Tr. sinuatum the thickness varies from 0.003 mm. on the dorsal side of the posterior sucker to 0.01 mm. on the dorsal side near the anterior sucker. It is also to be noted that the thickness is sometimes less on the dorsal than on the ventral side, eg. in Onchocotyle. Having described the investing membrane as it is according to my own observations, I may now refer to some views relating to its ’ STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 1l nature. These views may be divided into two classes, viz., that which regards the investing membrane as a true cuticle, and that which regards it as the transformed product of the originally cellular epidermis. Among the most recent writers on ectoparasitic Tre- matodes, Brandes represents the first view and Braun the second.” According to the former writer? the investing membrane is to be regarded as a true cuticle and as the secretion product of numerous unicellular glands scattered in groups in the cortical portion of the mesenchyma (Ectoparenchym), the presence of which he claims to have demonstrated in all the forms studied by him for the purpose. In opposition to this view, Braun” brings forward the fact that in Mono- stomum mutabile there are disseminated in the investing membrane (‘‘ Hautschicht’”) numerous oval nuclei with sharp outline and stain- ing weakly with picro-carmin. In absence of any direct embryolog- ical proof it is of course useless to dogmatise on either side. But, even laying aside the fact observed by Schwarze” that the investing membrane of Cercaria is a transformed epidermis as not quite conclu- sive with respect to that of the adult worm, the positive facts” at present known, when taken together seem to me to be strongly in favour of the view upheld by Braun. For instance, the observa- tion of Zeller® on Polystomum, that in it the epidermal cells are not cast off but have their nuclei merely shrivelled up, strongly points to 1). As stated in a previous note Monticelli is of the same opinion. 2). Brandes—l. c. 3). Braun—Ueber einige wenig bekannte resp. neue Trematoden. Verhandl. d. deutsch. zool. Gesellsch., 1892. p. 51. 4). Schwarze—Die postembryon. Entwickl. d. Trematoden. Zeitsch. f- wiss. Zoolog., Bd. 42, 1886. p. 49. 5). Itis perhaps hardly necessary to remark here that the “matrix cells” of Wierzejski are in reality the nuclei of the mesenchyma. Cf. Wierzejski in Zeitsch. f. wiss. Zoolog., Bd. 29, 1877. p. 552. 6). Zeller—Weiterer Beitrag z. Kenntniss d. Polystomen. Zeitsch. f. wiss. Zoolog., Bd. 27 1876. p. 262. 12 Ss. GOTO. the continued existence of the original epidermis. Besides, the differentiation of the investing membrane into the true cuticle and the subcuticle,' and the negative fact that, although I have directed my special attention to the point, I have utterly failed to observe those subcuticular glands so beautifully drawn by Brandes in his figures in the very same genera that he describes, strongly incline me to the view that the investing membrane of the ectoparasitic Trematodes is a transformed epidermis. I have indeed observed some cells in the ectoparenchyma which had the appearance of a gland (Pl. XXI, fig. 4), but I have not been able to find out any duct, and believe them to be cells.of the mesenchyma and will therefore describe them under that head. From his statements on p. 565 (op. cit.), I gather that Brandes regards the muscular fibres described by Poirier” in some species of Distomum as the ducts of the subcuticular glands; but Poirier’s figure in question is so clearly drawn that [ doubt whether one is justified to put another interpretation on it unless he has studied the very same species. In Brandes’ figures the muscular fibres and the ducts of the subcuticular glands are distinguished by different co- lours, but according to my own experience it is very doubtful whether such difference in colour reaction exists really in nature. Moreover the ducts of the subcuticular glands are drawn so fine in Brandes’ figures that one would be tempted to regard them also as muscular fibres if they were coloured alike; and for my own part I do so regard them. In this connection it may be mentioned that on examining once a series of sections of Hexacotyle grossa which were somewhat overstained, the terminal portions of the dorso-ventral muscular fibres were so deeply stained that they looked just like the efferent ducts of some 1). Poirier—Contribution 4 histoire naturelle des Trématodes. Archives de Zool. expér., 2. série., I. III, 1885. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 13 glands ; but on comparing it with another series of sections I was able to demonstrate clearly their muscular nature.” 3. Musculature. The musculature of the body of the Trematodes consists, as is well known, typically of four sets of fibres, viz., the circular, the diagonal, the longitudinal, and the dorso-ventral. ‘There are, however, some variations in different species and genera ; and I shall proceed to note them in the species studied by me. -In Microcotyle, Aine, Onchocotyle, Octocotyle, Monocotyle, Calicotyle, and Tristomum I find the musculature of the body to consist of the typical four sets of fibres ; but the different sets or Jayers are developed in different degrees in different species. Thus in Microcotyle, Axine, Octocotyle, Diclidophora, Monocotyle, and Calicotyle, the circular fibres are very fine and are directly applied to the basement membrane, so that they appear in sagittal sections of the worms as minute dots arranged at regular short intervals. In cross-sections of the worms they are very difficult to demonstrate. In Tristomum, on the contrary, the individual circular fibres are stronger and they are generally at some distance from the basement membrane (PI. XXIJ, fig. 4; Pl. XXII, figs. 4 and 7; Pl. XXIV, fig. 2), leaving a layer of mes- enchyma of variable thickness between. In Onchocotyle and Heaacotyle the circular fibres seem to be entirely wanting. In Dielidophora an additional layer of longitudinal fibres comes between the circular and the diagonal fibres (PI. X, fig. 4 & PI. XI, figs. 3 & 5). The individual fibres of this layer are separated from one another by an intervening mass of mesenchyma ; they are usually oval or circular in cross-section and are generally a little finer than those of the inner longitudinal layer. ‘This layer has also been 1). Compare on this question the more exhaustive discussions in Monticelli’s paper (Primo contributo etc., p. 202 et infra). : 14 S. GOTO. observed in some turbellarians. I have been able to demonstrate it in all the three species of Diclidophora which I have studied. The fibres are usually arranged in a single layer ; but on the ventral side of Diclid. tetrodonis they are, irregularly, more than one layer thick (PI. X, fig. 4). In the pedicels of the posterior suckers of the genus under consideration, only the circular fibres are present (Pl. XII, fig. 4); the longitudinal fibres forming in each an axial bundle which is attached to each sucker. The diagonal fibres that come next the circular, or in Diclidophora next the outer layer of longitudinal fibres, are most strongly developed in Tristomum. In Trist. ovale (Pl. XXIII, fig. 7 & Pl. XXIV, fig. 2) this layer consists of numerous fibres which are rather closely crowded and cross each other at variable angles according to the different states of contraction of the body. The individual fibres are somewhat weaker than the circular fibres. In Trist. sinuatum, on the other hand, the absolute number of fibres that constitute this layer is considerably less than in Tr. ovale, but the individual fibres are much stronger (PI. XXI, fig. 4), and cross each other at an acuter angle. ‘The last men- tioned fact can not only be demonstrated by an examination of prepara- tions in toto but is also evident from the fact that in sections of equal thickness of the two species a shorter portion of each fibre is cut in the one than in the other (cf. Pl. XXI, fig. 4 & Pl. XXIII, fig. 7). In Diclidophora the individual fibres of this layer are very much finer than the longitudinal fibres, but are comparatively numerous (Pl. XI, figs. 3& 6; PL X, fig. 4). In Hexacotyle, on the other hand, the fibres are not so numerous, but each one is only a little inferior in size to the longitudinal fibre. In Axine (PI. VIII, fig. 1) and Mono- cotyle (PI. XVIII, figs. 2 & 5) the fibres are very fine and not very numerous. Finally in Microcotyle the diagonal fibres are very weakly developed and can be demonstrated only in preparations in toto. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 15 As may be seen from the figures referred to, the layer of diagonal fibres is generally separated both from the circular and the longitudinal fibres by a thin intervening layer of mesenchyma. ‘The layer of inner longitudinal fibres is the most strongly developed of all the muscular layers of the body, both as a whole and in the strength of the individual fibres that constitute it. In Azine and in most species of Microcotyle, the fibres of this layer are not arranged in bundles but are almost uniformly scattered in a distinct layer of the mesenchyma (PI. IV, figs. 6, 7, & 8; Pl. VIII, fig. 1), and present oval or circular outlines in cross-sections. In Tristomum (Pl. XXIII, fig. 7 & Pl. XXIV, fig. 2), Onchocotyle (Pl. XV, fig. 10 & Pl. XVI, fig. 8), and Heaacotyle (Pl. XII, fig. 5) the fibres are united only into loose bundles ; but in nearly all the other species stud- ied by me, the fibres of this layer are associated in compact bundles, and present in cross-sections generally polygonal outlines, evidently due to mutual pressure. In some species, as in Hexacotyle acuta and Microcotyle reticulata, this inner layer of longitudinal fibres is again divisible into two layers. In the former species the outer of the two layers is constituted by a single layer of strong fibres at various distances from each other (Pl. XII, fig. 5). The longitudinal fibres of Microcotyle reticulata present some variations of arrangement, which will be briefly noted. In the more anterior part of the body (PI. V, fig. 6) the fibres are in this species distributed apparently without order, but the outer fibres are considerably smaller than the deeper ones. In the region of the vagina the fibres of the two layers are almost equal in size (PI. V, fig. 5) ; but in most portions of the body the longitudinal fibres are arranged in compact bundles, the outer fibres of which are considerably smaller than the inner (PI. IIT, fig. 4). In this species . the circular and diagonal fibres are but weakly developed and can not 16 8. GOTO, be demonstrated in cross-sections ; so that the layer of longitudinal fibres seem in such sections to be separated from the investing mem- brane only by a layer of mesenchyma (PI. III, fig. 4). . ‘The dorso-ventral muscular fibres are developed in very different degrees in different species, but they are, so far as I have observed, never wanting. They are but weakly developed in Axine and in most species of Microcotyle ; moderately in Diclidophora, Hexacotyle, Onchoco- tyle and Monocotyle ; and very strongly in Tristomum. It is, as is well known, the characteristic of the dorso-ventral fibres that they ramify into a number of fine branches towards their ends, and are inserted onto the investing membrane of the body. ‘They also traverse some internal organs, such as the testes and the vitellarium. In Monocotyle there are in the hindmost portion of the body an assemblage of striped muscular fibres (P]. XVII, fig. 5); but as these are present mainly in the posterior sucker, they will be described in that connection. 4, The Organs of Attachment. Under this head I include the suckers, both true and rudimen- tary, the glands, which subserve, according to my opinion, the same purpose, and the hooks, These, but especially the glands, are more numerous and of more varied structure than has hitherto been thought. Sucxers—The suckers may be classed, for the sake of description, into the anterior and the posterior suckers, of which the latter present more variety of structure than the former. The differences in both of them are characteristic of the genera, and so will be treated separately under each genus, except where another treatment is more desirable. Microcotyle and Axine—In these two genera the anterior and posterior suckers are of the same structure. They will therefore be treated of together and minor differences occasionally noted. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 17 The anterior suckers consist in both the genera of a pair of bag-shaped bodies with thick walls, and situated one on each side of the mouth-cavity (Pls. I, II, II, & VIII). The shape of the sucker is very variable according to the different states of contraction of the part of the body it belongs to, and also differs somewhat in different species ; but when in a state of rest it is generally circular, ellipsoidal, or egg-shaped (PI. III, fig. 8; Pl. VII, fig. 1). Its cavity is directly continuous with that of the mouth, and in most species of Microcotyle it is divided into two compartments by a struc- tureless, membranous septum that usually runs obliquely to the long axis of the body ; but the septum is also absent in some species, as in M. reticulata. Parona and Perugia” mention “ piccolissimi e splendenti corpicciuoli rotondi disposti a gruppi, od in una fascia” on the free margins of the suckers ; but I have not observed such struc- tures. ‘The wall of the sucker consists of very refractive, prismatic fibres of a yellowish colour traversing its whole thickness, which remain entirely unstained in haematoxylin, picro- or borax-carmin. These fibres are closely appressed to one another—the prismatic shape being apparently due to mutual pressure—and only a very thin layer of connective tissue is left between them. This layer of connect- ive tissue stains more or less, and in consequence, the substance of the wall of the sucker appears in sections striated at right angles to its surfaces. These fibres are different in appearance as well as in colour-reaction from the muscular fibres of the body, and are perfectly like those of the posterior suckers and of the mass of connective tissue around the terminal portion of the vas deferens in the genera 1). Parona e Perugia—Res ligustice, XIV. Contribuzione per una monografia del genere Microcotyle. p. 4. Estratto dagli Annali del Museo Civico di Storia Naturale di Genova. Ser. 2, vol. X, 1890. 18 8. GOTO. under consideration. In fact I consider them to’be more of an clastic than a contractile nature ; but the reasons for so regarding them will be stated at length further on under the head of general considerations (vide infra, p. 144). "The surface of the wall of the anterior sucker is covered on all sides by a thin cuticle (PI. III, fig. 8). On the side turned towards the mesenchyma, the fibrous connective tissue of the body forms, next to the cuticle, a thin dense layer similar in appearance to the basement membrane already described. Each anterior sucker is provided with two muscular bundles, one of which, the larger, is attached to its postero-lateral part and the other, the smaller, to the postero-median part (PI. III, fig. 8). Posteriorly these muscular fibres become’ mingled with the longi- tudinal fibres of the body. , The physiology of suction will be considered later on (vide p. 147). Posterior suckers—These are usually very numerous in both Azine and Microcotyle, and in most species of the latter are symmetrically arranged on both sides of the. caudal disc, while in the former genus they are always asymmetrically arranged on the two sides in ac- cordance. with the general asymmetrical form of the body already described. ‘Their sizes differ in both the genera in different parts of the caudal disc; the general rule being that they are largest at the middle of the disc and diminish in size towards the ends. The pos- terior suckers are however mostly smaller than the anterior ones. In M. reticulata and M. sciaenae the suckers are more numerous on the right than on the left side ; and in the latter species the eandl disc is bent at an angle towards one side, while in the former the right side presents only a greater curvature. This asymmetry of the body is, as already stated, due to the fact that its one side is longer STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 19 than the other. In J. sciaenae the suckers of the two sides do not present any considerable difference in size”; bat in Jf. reticulata those of the left side are much larger and fewer in number than those of the other. Thus, in the latter species a measurement of the breadths of the corresponding suckers of the two sides at about the middle of the caudal disc gave for the left side 0.227 mm. and for the right 0.145 mm. In the specimens of the two species figured on the plates the right side is longer and bears more suckers” than the left ; but whether this is constantly the case or not I have not had a sufficient number of specimens to decide (of. infra de Axine). In M. caudata, M. sebastis, M. elegans, and M. fusiformis the caudal disc extends for some distance anteriorly from the point where it becomes continuous with the body proper ; but in all the other species it does not, and the suckers are arranged merely along the lateral margins of the body. In Amine the asymmetry which we have observed in the two species of Microcotyle above mentioned is carried one step farther, and one of the sides of the caudal disc makes an angle with the corres- ponding side of the body proper (PI. VII), and appears like the posterior border (the “appendice ptéroide” of v. Beneden). The other side bears only a small number of suckers. In some species, as in A. triangularis (PI. VII, fig. 7), the suckers of the two sides are nearly of the same size; but in A. heterocerca. (PI. VII, fig. 1) they are of very different sizes on the two sides. Thus in one specimen, one of the largest suckers on the longer side was 0.60 mm. in breadth, 1). For minuter details see description of species. 2). Parona and Perugia (Res ligusticae, XIV, p, 38) believe that a similar asymmetry occurs in DM. erythrint; but in my opinion, the supposed asymmetry in this case is only apparent, having been caused by pressure and the twisting of the body at the point of attachment of the caudal disc, as occurs very often when the worm.is observed under the cover-slip. The real asymmetry could only be caused by the unequal length of the sucker- pearing portion of the two sides, and there is none in the species in question. 90 8. GOTO. while that of the other was only 0.095 mm. On the longer side the largest suckers are, as.in Microcotyle, generally found in the middle portion ; but on the shorter side the suckers gradually diminish in size from before backwards. In A. aberrans, however, the suckers are all of nearly equal size on the longer side. Lorenz” and others mention the suckers only on the longer side; but in all the species studied by me they are present on both sides (Pl. VI, figs. 1, 5, & 7). Now as to structure, each posterior sucker may be likened to a flattened, rectangular bag open on the ventral side, the broader sides of which face forwards and backwards and have a very thick wall, while on the narrower sides the wall is very thin. This bag is supported by a chitinous” frame consisting of five pieces, viz., four lateral pieces in pairs and one median. Of the four lateral pieces two are imbedded in the substance of the anterior wall and two in that of the posterior wall, both along the lateral margins (PI. II, fig. 7). These are curved like a hook, are somewhat triangular in cross-section, and are entirely solid. ‘The rods of the anterior wall (a) are, at the bottom of the sucker, directly continued on into the posterior wall (a’); and the whole has, therefore, somewhat the form of a hook (cf. PI. III, fig. 1 & Pl. II, fig. 7). Those of the posterior wall (b) end bluntly at the 1). Lorenz—Ueber die Organisation der Gattungen Axine und Microcotyle. Wiener Arbeiten von Claus. Bd. I, Hft. 3, 1878. p. 4. 2). I have used the word “chitinous” here and shall use it in describing the hooks in accordance with the usual custom; but it should be noted that these pieces are sometimes well stained with haematoxylin—exactly under what circumstances I have not been able to make out; but one condition seems to be that the specimen be preserved after a certain dis- integration of the tissues has set in—and that the hooks are soluble in a solution (35 °%o) of caustic potash. 3). I take this opportunity to correct my statement on this head with regard to Diplozoon Nipponicum. In my paper on this worm (i.c.) I have described the piece imbedded: in the posterior wall and projecting towards the median piece as the process of the paired (lateral) piece of the posterior wall, whereas it is in reality the direct continuation of the paired piece of the anterior wall, just as in Microcotyle. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. - 9} bottom of the suckers. ‘The median piece (c) is also imbedded in the substance of the wall, and has the form of a regular U, one arm of which lies in the anterior and the other in the posterior wall. The two ends are however somewhat different in form (PI. III, fig. 1). In the anterior wall the median piece ends with two spine-like processes diverging from each other and almost meeting the ends of the lateral pieces. In the posterior wall, on the contrary, these processes are very much shorter and blunter, and the piece bears between them a dagger-shaped, hollow, terminal piece (PI. III, fig. 1). In cross- section, the median piece presents for the greater part of its whole length, a rectangular outline, and is seen to be hollow. . A little above the bottom of the sucker its cross-section presents the appearance drawn in fig. 7 (c on the upper side), PI. IT. In Axine the number and relative position of the chitinous pieces are just the same as in Microcotyle (Pl. VII, fig. 2). The internal hollow of the median piece is, however, divided in A. heterocerca into numerous compartments by thin septa. This piece presents also a termination in the posterior wall which is somewhat different from that in Microcotyle and is figured in fig..2 u, Pl. VIL. The anterior and posterior walls of the posterior sucker are very thick and are directly continuous with.each other at the bottom of the sucker. Their substance consists of numerous prismatic, very refract- ive fibres closely appressed to one another and leaving a very thin, deeply stained layer of connective tissue between. These fibres are to all intents and purposes exactly similar to those of the anterior suckers (PI. II, fig. 7 & Pl. VII, fig. 4). The wall is entirely surrounded on all sides by a thin cuticular membrane. In Azine, however, the cuticle.of the internal surface is chitinized in parallel zones running parallel to the slit-like mouth of the sucker (PI. VII, fig. 4). The Jateral walls of the posterior suckers are very thin and membranous. 29 8. GOTO. The muscular fibres of the posterior suckers are attached to the median chitinous piece of the posterior wall near the bottom of the suckers (Pl. II, fig. 7). They are divided into a few bundles, and are directly continuous with the longitudinal fibres of the body. Octocotyle, Diclidophora, and Hexacotyle—The anterior suckers of these three genera present nothing specially different from those of Awine and Microcotyle, and I shall therefore pass them over, merely referring the reader to the figures (PI. IX, fig. 8; PI. X, fig. 6; Pl. XIII, fig. 1 & 4). I shall, however, note that in all these, there is no membranous septum, that in Dichidophora (PI. X, fig. 6) some of the fibres of the internal bundle of muscle of one ‘side cross over to the sucker of the other side, and lastly that in Hesxacotyle the suckers are exceedingly small. ‘The posterior suckers are, however, very different in the three genera, so that they will be treated separately. Posterior suckers—In Octocotyle the four pairs of. posterior suckers are arranged merely along the ventro-lateral border of the hind- most part of the body, so that the caudal disc is, as already mentioned, directly continuous with the body proper. Each sucker is somewhat bean-shaped, and is raised a little above the general surface of the ventral side (PI. IX, figs. 1, 4, & 7). The anterior and posterior walls are very’ thick, ‘and consist of prismatic fibres just as in Microco- tyle and Axine. The. chitinous framework consists of a simple basal piece (PI. IX, figs. 3 & 10, a) anda pair of marginal pieces (b), both imbedded in ‘the substance of the wall. The basal piece is nearly straight and bears at each end a short process which ‘makes a right angle with it, and is directed towards the mouth of the sucker. The marginal pieces are U-shaped and are imbedded in the maleate of the anterior and posterior walls along their very margins, so that each STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 93 piece, together with the basal piece, encloses the wall like a picture in its frame, The muscular fibres are attached to the basal piece, and are direct continuations of the longitudinal fibres of the body. In O. major the marginal chitinous piece is longer than in O. minor. In Diclidophora (P1. X) the four pairs of suckers are, as already stated, borne, in most species, each at the top of a peditel; but in Diclid. sessilis the stalk is very short, and in Diclid. tetrodonis entirely wanting. But that these species are to be included in the same genus will, I believe, be scarcely contested by any one who has studied their anatomy. ‘The suckers are ina surface view usually circular, as in Diclid. sessilis and Diclid. elongata, but are sometimes slightly elliptical as in: Diclid. tetrodonis, and are arranged in a semicircle or horse-shoe shape. ach sucker is as a whole nearly hemispherical, and its open end is directed obliquely towards the ventral side. The wall consists of prismatic fibres, and is divided into four equal parts by as many chitinous rods radiating from the centre of the sucker (Pl. XII, fig. 4). The whole number of the chitinous pieces is eight for each sucker, and are represented in their natural positions as seen from. the surface in fig. 1, PI. XII. A T-shaped piece (a) is situated with the point of junction of the leg and arms exactly in the centre, with the two arms of unequal length extending some way towards the periphery of the sucker, and. the leg extending quite to the periphery and here dividing into two branches which diverge from each other in a diame- trically opposite direction. The portion of this chitinous piece corres- ponding to the leg of the T is hollow, but the remaining portions are all entirely solid. Another hollow piece (b) extends from the centre in an opposite direction from the leg of the T-shaped piece towards the periphery, which it, however, does not reach. ‘The hook- like piece marked ¢ is paired, and bears at its angle a process which is seen in sections to lie outside the substance of the fibrous wall (PI. Q4 S. GOTO. XI, fig. 4, ¢'). Somewhat similar but shorter pieces (d) lie in pairs at the peripheral end of b. A short spine-like piece (e) lies opposite the process of the piece ¢ with its pointed end directed towards the piece d. These last-mentioned pieces, the pieces d, and the peripheral portions of the pieces a and ¢ are imbedded at the very margin of the fibrous wall of the sucker (PI. XII, fig. 4). ‘In the natural position of the suckers these chitinous pieces are, in most’ species, placed in such a way that the leg of the T-shaped piece is directed towards the median line of the body in the anterior’ suckers, while in the posterior suckers. it is directed obliquely. forward (Pl. X, fig. 5). In Dichid. tetrodonis, however, the same piece is directed forwards in the posterior three pairs, but backwards in the foremost pair. This difference of arrangement of the pieces of the chitinous framework in the foremost pair of suckers is a curious exception to the rule. Each of the four quadrants into which the wall of the sucker is divided, presents in radial sections an oval or semicircular mass, con- sisting, as in the species already described, of prismatic fibres travers- ing its whole thickness at right angles to its free surface (PI. XII, fig. 4). The four quadrants are separated from each other in the centre by an intervening mass of connective tissue, in which is imbedded the central part of the T-shaped piece (a) already described. The free surface of each quadrant is covered with a thick cuticle which is locally chitinized in short staff-shaped pieces arranged in series. ach series of these staff-shaped chitinous portions is, in a surface view, roughly speaking somewhat semicircular (PI. XII, fig. 2), with the open side directed towards the centre of the sucker. In fig. 1, Pl. XII, the series have been a little displaced by pressure. To- wards the mesenchyma the wall of the sucker is bounded by a thin, cuticular membrane. Around the outer margin of the sucker the STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 95 investing membrane of the body makes a fold filled with connective tissue within, and forms a marginal membrane (PI. XII, fig. 4). Strong bundles of muscular fibres are attached to each sucker. The principal bundles are, as may be seen in fig. 4, Pl. XII, three in number, and are attached one to the inner margin of an inner quad- rant, another to its outer corner, and the third to the central part of the T-shaped chitinous piece. The individual fibres that constitute these bundles are very large. The larger part of the bundles of each sucker are merged in a single bundle at a short distance from it, and the bundle thus formed traverses the axis of the pedicel of the sucker, when such is present ; but a small portion forms one or more bundles which connect the suckers with one another. I have represented this connection among the suckers as I have observed it in Diclid. sessilis in fig. 5, Pl. X. In Hexacotyle, there are four pairs of posterior suckers, and these are arranged on each side of the median line in a line parallel with the posterior border of the body. ‘The median pair of all is, however, very much smaller than the others, and has been entirely overlooked by some naturalists ; but it has exactly the same structure as the other pairs. Hach sucker is shaped like a thick but very shallow, elliptical saucer (PI. XII, fig. 7), with its longer axis directed at a right angle to the posterior border of the body, and is provided with three chitinous pieces imbedded in the substance of its wall, which consists, as in the cases already described, of prismatic, refractive fibres closely pressed against one another. The positions of these chitinous pieces are represented in figs. 1 & 4, Pl. XIII; and in figs. 3 & 6 each piece has been separately drawn, so that any detailed description of them is, I believe, unnecessary. It is to be remarked, however, that in both the figures, a represents the piece at the anterior end of the sucker, ¢ that at the posterior end, and b the central piece, and that 26 8. GOTO. this last piece is imbedded in the wall of the sucker with its longer axis coinciding with the minor axis of the sucker. The substance of the wall is-bounded, as in other genera, both towards the mesenchyma and the exterior by a cuticular membrane, except in the central part where the chitinous piece is imbedded, and where this is in direct contact with the mesenchyma. A strong bundle of muscular fibres, which are direct continuations of the longitudinal fibres of the body, are attached to each of the chitinous pieces of the sucker (PJ. XII, fig. 7). Onchocotyle—The position and structure of the suckers are somewhat peculiar in this genus. Anterior sucker—An anterior sucker is distinctly present, although some writers have denied the fact, and is situated around the mouth- cavity, and presents some resemblances of structure. to that of the distomes. In cross-section, it is elliptical in outline, and is seen to occupy nearly the whole of this region of the body, merely leaving a smal] portion of mesenchyma in the lateral parts (Pl. XV, fig. 4). ‘The ventral half is much smaller than the dorsal half, just as in the distomes (Pl. XV, fig. 3). The substance of the sucker consists of connective tissue interspersed with nuclei, some of which are surrounded by scanty masses of granular protoplasm, and of muscular fibres, most of which are very fine, and, radially traversing the con- nective tissue, are attached to the basement membrane which separates the substance of the sucker from the investing membrane of the body. Besides these radial fibres, strong muscular fibres which act as sphincters are present in the foremost part of the dorsal half of the sucker (Pl. XV, fig. 3, sph.) Posterior suckers—These are eight in number arranged in four pairs, one of which is, however, very different both in structure and position from the others. The latter are arranged in a_horse- STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 97 shoe shape seemingly on the dorsal but morphologically on the ventral side of the body, and are hemispherical in shape. The wall consists, as in the preceding cases, of prismatic, refractive fibres set closely against one another, and is bounded on all sides by a cuticular mem- brane. This membrane is chitinously thickened into band-shaped strips in an antero-posterior direction (Pl. XV, fig. 8). The wall of each sucker is again divided in two, each half crescent- shaped in cross-section, and separated from the other by an in- tervening zone of mesenchyma, in which a single, strong, hollow, chitinous piece of semicircular shape and provided with a hook at one end lies imbedded (Pl. XV, figs. 8 & 9). The inner half of the wall is smaller than the other. ‘There is no distinct marginal mem- brane, but the investing membrane presents a fold all round at a short distance from the mouth of the sucker (fig. 8). In the natural position, the hooked end of the semicircular chitinous piece is directed towards the anterior end of the body ; and a ‘strong bundle of muscular fibres is attached to the other end of the piece. Besides this, detached fibres are also attached to the lateral margin of the suckers. The remaining pair of suckers, which are much smaller,: is situated at the extremities of the bifurcations of the caudal appendage which constitutes such a salient feature of this genus. Each sucker has the shape of a water-melon with a constriction at the posterior part ; and its mouth is very narrow (PI. XV, fig. 2). Its cavity is also very narrow ; but the wall is very thick, and consists entirely of a connective tissue interspersed with nuclei, which are more abundant in the posterior portion of the sucker (PI. XV, figs. 1 and 2). The fibres of the connective tissue are mainly arranged radially, and form an irregular network. The wall is bounded towards the mesenchyma by a thin, cuticular membrane, and its inner surface is lined by the 98 8S. GOTO. direct continuation of the investing membrane of the body. The pair of suckers above described has been mistaken by Tas- chenberg” for the terminal vesicles of the excretory system (see p. 5). Calicotyle—Anterior sucker—In this genus the anterior sucker is developed far more imperfectly than in those hitherto described, and is also totally different from them in structure. It is constituted by a deep invagination of the ventral side of the body just behind the mouth (Pl. XIX, fig. 8). From the inner surface of the investing - membrane that lines the cavity of the sucker, numerous muscular fibres take their origin, and diverging in various directions, become continuous with the longitudinal fibres of the body. Besides these, circular fibres are exceptionally developed in this part. Owing to this local development of the musculature, the anterior sucker seems in a surface view to be pretty well defined from the surrounding mes- enchyma, and to form a distinct organ (Pl. XIX, figs. 1 and 4); but sections show that this is by no means the case. Morphologically speaking, it would be more correct, therefore, to call it a rudimentary sucker (pseudoventose). Posterior sucker—This is situated on the ventral side at the pos- terior extremity of the body. Its ventral surface is marked out into a central heptagonal and seven peripheral areas by elevations of the ventral surface radiating from the wall of the central polygon. ‘This central polygon does not occupy exactly the middle of the sucker, but is situated a little more anteriorly. The hindmost area which is the largest, occupies the median line of the body, while the others are arranged symmetrically on either side, and decrease in size from behind forwards (PI. XIX, figs. 1 and 3). The musculature of the sucker consists of three sets of fibres, which may be called the radial, the transverse or dorso-ventral, 1). Taschenberg—Weitere Beitriige, p. 13. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 99 and the circular. The radial fibres must again be divided into two groups. In one they are direct continuations of the longitudinal fibres of the body, and on entering the sucker diverge irregularly more or less in all directions, and are inserted, some to the elevation that forms the wall of the central polygon, others to the periphery of the sucker. The other group of radial fibres includes those fibres which start from the central polygon, and passing along the radial elevations, are inserted in the periphery of the sucker. The transverse or dorso-ventral fibres merely traverse the thickness of the sucker, and are inserted into the basement membrane. Like the dorso-ventral fibres of the other parts of the body, they ramify into a number of small branches towards their ends. The circular fibres are present only on the dorsal side; they are arranged in circles concentric with the circumference of the sucker, and are comparatively few. ‘lhe inter- spaces between the muscular fibres are filled by 2 connective tissue similar to that of the body (Pl. XIX, fig. 6). . Monocotyle—Morphologically speaking there is no anterior sucker in this genus, but physiologically speaking there is. The dorso-ventral as well as the circular fibres of the body, namely, are strongly developed at the anterior end around the mouth Chi. 2 V Lo, fig. 4), so that the arrangement of muscular fibres is very similar to that which obtains in the anterior suckers of Onchocotyle and of the distomes. ‘There is even a sort of marginal membrane on the ventral lip (PL XVIII, fig. 4, a). That this anterior portion acts as a sucker is beyond doubt; for I have observed the worm execute an active leech- like locomotion by alternately attaching and detaching the mouth and the posterior sucker. In fact, the structure here described is a prelim- inary step to the formation of such anterior sucker as that of Onchocotyle and the distomes ; the only difference being that in the latter, the sucker has been distinctly separated from the surrounding 30 8. GOTO. mesenchyma by a compact membrane of the nature of connective tissue. Though closely allied to Calicotyle, the present genus presents some peculiar feature in the structure of the posterior sucker. Posterior sucker—This has exactly the shape of a circular saucer, and is attached to the body by a short stalk. It is provided with a “marginal membrane all round, which is exactly of the same nature as that of the posterior suckers of Diclidophora ; only its connective tissue is more compact (Pl. XVII, figs. 2, 4, and 5). A further difference is that the membrane bears on its ventral surface numerous chitinous projections arranged in radial series ; there being five of these chitinous projections in each series (figs, 2,4, and 5). The internal surface of the sucker is divided into eight equal secants by as many radiating elevations, the positions of which are such that two of them coincide with the median line of the body. ‘These radial elevations also bear on their surfaces each a series of chitinous pieces which are figured in radial section in fig. 8 (Pl. XVII), and in tangential section in fig. 8. In the latter figure it appears as if these pieces have been formed simply by local chitinisations of the investing membrane of the body. These so-called chitinous pieces are deeply coloured by haematoxylin, more deeply than the investing membrane. The muscular fibres of the sucker are very different in structure from those hitherto described, and are all arranged, with an insigni- ficant exception, dorso-ventrally (Pl. XVII, figs.4, 5, 7,and 12). They are striped, and a single fibre traverses ‘the whole thickness of the sucker. ach fibre (about 0.01 mm. thick) consists of very refractive, narrow zones, which deeply stain in haematoxylin, alternating with broad, non-refractive zones, which are but slightly stained with haematoxylin and are finely but distinctly striated longitudinally. Each of these non-refractive zones is again crossed at its middle by a STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 31 i very narrow, secondary zone of much refractive power and well stained in haematoxylin. ‘This zone is so narrow that it appears only as a fine line, and its refractiveness and capacity of being stained with haematoxylin are much inferior to those of the primary zone. I have observed some of these fibres bifurcating towards one of their ends (Pl. XVII, figs. 7 and 12). The striped fibres above described are almost uniformly dis- tributed in the secants into which the sucker is divided, as well as in the radial elevations themselves. In the latter, the fibres lie at right angles to their length, .and are consequently very short (Pl. XVII, fiz. 8). ‘They may also be entirely absent from them for a more or less wide extent (PI. XVII, fig. 3). On the other hand, they are wholly absent from those parts of the sucker which lie below the radial eleva- tions. Here the substance of the sucker is entirely formed of connect- ive tissue, the fibres of which unite into bundles on the ventral side in such a way as to form a series of window-like cavities (PI. XVII, fig. 3) ; while on the dorsal side the fibres form generally a compact network, leaving only here and there a number of large cavities. The centre of the sucker is wholly devoid of striped fibres, and is sharply defined from the surrounding parts by a membrane of connective tissue (PI. X VIL, figs. 4 and 5). This central part is traversed by the terminal portions of some of the longitudinal fibres of the body, which are here formed into bundles and are inserted, some into the very centre of the sucker, others more peripherally into the membrane of connective tissue that separates the centre of the sucker from the surrounding parts. As may be inferred from the above description, the sucker of Monocotyle is divided both externally and internally into eight equal secants—externally by radial elevations, and internally by correspond- ing radiating septa formed of fibrous connective tissue and wholly destitute of striped muscular fibres. 32 8. GOTO. As has already been mentioned, the striped fibres are also present in the posterior part of the body (Pl. XVII, fig. 5), and this portion is, as in the sucker, sharply separated from the surrounding part by a distinct membrane of connective tissue (figs. 5 and 12). The striped muscular fibres have hitherto been observed, so far as I know, only in the sheath of the proboscis of Tetrarhynchus» among the Plathelminthes. Tristomum—As is well known, the anterior suckers are, in this genus, situated in a pair near the anterior extremity of the body, on both sides of the mouth or a little before it. They are more or less circular, sometimes perfectly so and sometimes a little elliptical. Internally the substance of the suckers is directly continuous with the mesenchyma of the body, and the suckers are therefore to be regarded. merely as local expansions of the body. This view is strengthened by the fact that the muscular fibres of the suckers are the direct continua- tions of those of the body. The dorso-ventral fibres are arranged exactly as in the body, with the only difference that they are here more strongly developed in accordance with their task as the chief agent of suction (PI. XXII, fig. 5). The longitudinal and diagonal fibres, on entering the suckers, become more or less intermingled with each other ; but the former mostly radiate straight towards the periphery, while the latter curve round, and are mostly arranged more or less concentrically with the circumference of the suckers ; the terminations of both the groups of fibres being attached to the investing membrane on the dorsal as well as on the ventral side. The circular fibres of the body also enter the suckers ; but they here lose their typical arrange- ment, and become intermingled with the diagonal fibres. 1). Pintner—Untersuchungen ii. d. Bau des Bandwurmkoérpers. Wiener Arbeiten, Bd. 3, 2. Hft., 1880. p. 50. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 33 In most species of this genus described in this paper, the invest- ing membrane of the suckers is exactly similar to that of the other parts of the body ; but in Trist. ovale the ventral (concave) surface is raised into numerous conical papillae, at, the top of which open: the efferent ducts of the numerous unicellular glands afterwards to be de- scribed. Along the inner border of the suckers also there are, in this species, numerous larger papillae crowded together (P]. XXIII, fig. 5). Posterior sucker—The posterior sucker of Tristomum is more or less hemispherical in form, and is very similar in structure to that of Calicotyle. In most species, however, a short stalk may be dis- tinguished (Pl. XX, fig.5). The ventral (concave) surface is divided, as in Calicotyle, by radiating elevations into a central and seven peripheral polygons that surround the former ; the hindmost of the peripheral polygons invariably occupying the median line. In some species (Tr. sinuatum and Tr. rotundum), the central area forms a regular heptagon ; but in the majority of the species studied by me, its form is that of a heptagon to one side of which an isosceles trapezoid has been apposed by the shorter one of the two parallel sides and with the boundary line between the two blotted out (Tr. ovale, Tr. foliaceum, Tr. Nozawe, Tr. biparasiticum). In Tr. sinuatum the two radial sides of the hindmost peripheral area bifurcate near their outer ends and there form with the margin of the sucker each a small triangle (PI. AX, fig. 1). A marginal membrane is always present, and is thrown into wavy folds; so that in a surface view, it seems as if it consisted of numerous rectangular portions. Its investing membrane is very thin, and its substance consists of a syncytium with its nuclei irregularly scattered, and traversed by numerous fine, transverse, muscular fibres (Pl. XXII, fig. 3). According to my observations on T'r. sinuatum, the species to 34 S. GOTO. which I have mainly directed my attention in this respect, the arrangement of the muscular fibres of the posterior sucker is somewhat different from that described by Niemic.” The fibres may be dis- tinguished into four groups. (1) The radial fibres. These must again be distinguished, as in Calicotyle, into two subgroups : (a) those that are direct continuations of the longitudinal fibres of the body, and entering the sucker, diverge irregularly in all directions, and are inserted to the periphery of the sucker. They moreover cross each other in the stalk of the sucker; those coming from the two sides of the body going to the opposite sides of the sucker (Pl. XX, fig. 5). (b) Those fibres that are confined to the radiating spokes. These start from the margins of the central polygon, and running in small bundles along the spokes just under the investing membrane, are inserted, likewise, in the periphery of the sucker (Pl. XX, fig. 7, rad. mus. b.). (2) The circular fibres, i. ¢., those that run concentrically with the circumference of the sucker. This group must also be divided into two sub-groups. (a) Those that belong to the ventral side. These are arranged in two sets, an outer and an inner. The fibres of the outer set are arranged ina single layer ; while those of the inner set are united into strong bundles, which are separated from one another by the transverse fibres to be described presently, and from the outer set by an intervening layer of mesenchyma (PI. XX, fig. 6, cir. mus.) The circular fibres of the dorsal side are arranged in a single layer, and occupy an exactly similar position to that of the outer set of the ventral side. (3) The transverse fibres, i. ¢., those that traverse the thickness of the sucker, and run therefore at right angles to its two surfaces. They are, in Tr. sinuatum, very strongly developed, and, like 1). Niemic—Recherches morphologiques sur les ventouses dans le régne animal. Recueil zoolog. suisse. 'T. II, 1885. I have not been able to gain access to this paper, and am indebted for its account to Braun’s “ Wiirmer ” in Bronn’s “ Klassen u. Ordnungen ” and to Monticelli’s “ Saggio di una morfologia dei Trematodi.” STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 35 the dorso-ventral fibres of the body, ramify towards their ends (PI. XX, figs. 5, 6, and 7). (4) The last group of fibres is confined to the radial elevations, and simply traverse these from side to side, just as the transverse fibres do the sucker (Pl. XX, figs. 5 and 6). These fibres are absent where three elevations unite (PI. XX, fig. 4). Epibdella—I have been able to obtain but few specimens of this genus, and these were so badly preserved that I have not been able to make out the minute anatomy of the various parts; but a few points will be noted. The anterior suckers occupy nearly the same position in the body as in T'ristomum. They are, however, not so well developed as in that genus, and are more of a membranous character. The two suckers are also united with each other by a membranous anterior portion of the body that lies between them. The posterior sucker is either circular or elliptical in outline, and is provided with a marginal membrane. ‘There is no division of the ventral surface into separate areas as in Tristomum. The musculature consists of the cireular (in two sets, dorsal and ventral), the trans- verse, and the irregular radial fibres which are direct continuations of the longitudinal fibres of the body ; and these various fibres are arranged exactly like those of the same name in the posterior sucker of Tristomum. Sticky Guanps—Under this head I include all those glands which are present in various parts of the body and are not intimately connected either with the genital organs or the digestive system. They all open on the free surface of the body, and some of them are closely connected with the suckers. Microcotyle and Octocotyle—In all the species of these two genera that I have studied, except M. reticulata, there are three groups of sticky glands in the anterior part of the body in front of the 36 8. GOTO. anterior suckers. One of them lies in the median line near the apex of the body, while the other two are situated behind it in a pair, so that the three occupy the apices of an isosceles triangle, the base of which is perpendicular to the long axis of the body. In some species, however, they are nearly in the same straight line (Pls. Iand II). In section, each group is seen to consist of a limited number (5—7) of goblet- shaped bodies with a thin wall, and filled with numerous refringent granules of a yellowish colour, which do not stain at all in borax- carmin and only slightly in haematoxylin. It is, however, to be noted that their affinity for the latter dye varies according to dif- ferent stages of secretory activity. The necks of the goblet-shaped bodies of each group open together on the ventral side (PI. III, fig. 7). I have not been able to observe any nuclei in them; but I believe they are to be regarded as highly modified cells, and each group, there- fore, as an assemblage of unicellular glands. | These glands have been observed by previous writers, but their nature has been variously misunderstood. Wan Beneden and Hesse,” for instance, describe them as being “ destinées & remplir les fonctions de machoires.” Lorenz” on the other hand, considers them as tactile organs. Parona and Perugia® are of the same view, ‘“‘seouendo in questo il modo di pensare del Diesing e non quello di ? v. Beneden ed Hesse.” Now, the bodies in question are not situated in the cavity of the mouth as v. Beneden and Hesse think, and their glandular nature is, I believe, evident from their structure above _ described. 1). V. Beneden et Hesse—Recherches sur les Bdellodes et les Trématodes marins, 1863. p. 113. ; 2). Lorenz—Ueber die Organisation der Gattungen Axine u. Microcotyle. Wiener Arbeiten, Bd. I, Hit. 3, 1878. p. 24. : 3). Parona e Perugia—Res ligusticae, XIV.—Annali del Museo Civico di Storia Naturale di Genova, Ser. 2, Vol. X, 1890. Estratto p. 4. STUDIES ON THE ECIOPARASITIC TREMATODES OF JAPAN. 37 Axine, Diclidophora, & Mic. reticuiata—In these forms, there is only one pair of sticky glands. They are situated just in front of the mouth (PIL. I, fig. 5; Pl. VIL). Their torm differs much from that described in Microcotyle and Octocotyle and approaches more that of the sticky gland described by me in Diplozoon Nipponicum.? They are constituted, namely, by an invagination and transformation of the investing membrane of the body. They are shown in section in fig. 3, Pl. III. As may be seen from it, the cavity of the invagination is lined by a granular mass, which is a direct continuation of the juvesting membrane. The granules are very similar in appearance and in their reaction towards staining fiuids to those of the sticky glands of Microcotyle. Here, moreover, there is no distinct basement _ membrane but instead merely a dense layer of “connective tissue gradually passing into the surrounding mesenchyma. ‘There are in M. reticulata also some muscular fibres that come from the dorsal side of the body and terminate in the dense layer of connective tissue just mentioned (PI. III, fig. 3). Calicotyle—In this genus, there are two pairs of sticky glands at the anterior end of the body, the efferent ducts of which are com- pletely separate one from the other, but open very close to each other near the anterior apex of the body. Hach gland is a goblet-shaped, hollow body situated on the external side of the pharynx, with a long neck (the efferent duct) which proceeds towards the anterior end of the body, where it opens. One of the pairs has a little shorter neck than the other. ‘The goblet-shaped portion has a thick wall consisting of a granular substance, which, in the specimen I have examined, is stained well in borax-carmin ; but as I have been able to obtain only a single specimen of this genus, I can not state whether it is always so or not. The wall of the efferent duct consists of a thick membrane, 1). Goto—On Diplozoon Nipponicum, n. sp. This Journal, vol. IV, Pt. I, 1890. p. 166. 38 8. GOTO. on the inner surface of which some granulations are to be observed (Pl. XIX, figs. 1 and 4). As to the cellular structure of this gland, which I believe is here described for the first time in Calicotyle, there are, I think, two alternative views to be considered. One would be to regard it as being multicellular in its origin and to have been formed by the in- vagination of the epidermis; the other would be to regard it as unicellular. Against the second view it may be urged that these glands are too large to be regarded as unicellular—incomparably larger than the unicellular glands of similar function and position of the Gyrodactylidae—and that the presence of the internal cavity strongly points to the other view ; while the similarity of its wall to that of the similar glands of Diplozoon, Axine, and Diclidophora is also in favour of the first view. Monocotyle—In this genus there are four pairs of sticky glands at the anterior end of the body, and one in the posterior part. ‘The anterior glands are of exactly the same structure as those of Microcotyle, and are arranged at equal distances along the front part of the’ lateral border of the body, on both sides of the anterior: notch (PI. XVII, fig. 1). The posterior sticky glands are situated in that part of the body which lies dorsally to the posterior sucker, and consist of numerous small cells, each with a distinct, deeply stained nucleus and a very finely granular or almost homogeneous cytoplasm which seems to be destitute of an external membrane. The cells are of various sizes, apparently owing to the different stages of secretory activity ; and the smaller ones are more deeply stained and have more finely granular cytoplasm than the larger ones (Pl. XVII, figs. 5 and 12). In some parts, these cells are so closely pressed against one another that their boundaries become indistinct, and the whole appears somewhat like a STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 39 syncytium (Pl. XVII, fig. 6). They are distributed, both on the ventral and dorsal sides (PI. XVII, fig. 5), into two groups, the right and the left ; and the cells of each group discharge their products into a common large duct” provided with a distinct membrane, which opens by a minute aperture on the ventral surface of the lateral portion of the body, at a short distance from the stalk of the sucker (PI. XVII, fig. 12, gi’). The secretion of these granular cells is a very granular fluid, the granules staining very deeply in haematoxylin. Tristomum and Epibdella—As has already been observed by Monticelli,” there are, in these genera, numerous unicellular glands in the anterior portion of the body ; and in some species of T'ristomum (T. sinuatum and T. biparasiticum) there is in addition a series of groups of unicellular glands which open at regular intervals along the lateral borders of the body. The cells of the anterior sticky glands are more or less goblet- shaped, with a very thin membrane, a more or less granular cyto- plasm, a nucleus with-a single nucleolus, and a long neck. The nucleus sometimes occupies the central position, but is sometimes quite peripheral. The cells are distributed almost uniformly, or in groups separated from one another by dorso-ventral muscular fibres, in the mesenchyma of the suckers and of the anterior, median portion of the body. In some species of Tristomum (as in T. ovale) the necks of the gland-cells open at various points on the ventral surface of the sucker, at the tops of small, conical papillae (Pl. XXII, fig. 5), and especially along the inner border of the sucker, where the papillae are, as already mentioned, larger and very numerous (fig. 5). In T. sinuatum the glands open mainly along the margin of the anterior 1). Its size, however, probably varies according to the quantity of its content. 2). Monticelli—Di aleuni organi di tatto nei Tristomidi: Bollet. della Soc. di Naturalisti in Napoli, Ser. 1, Anno 5, vol. V., 1891, fas. 2. 40 8. GOTO. sucker, on the ventral side, where the basement membrane is some- what indistinct ; but they probably open also on the whole ventral surface, although I can not make a positive statement to that effect (PI. XX, fig. 10). In Epibdella, on the other hand, the glands open on the whole ventral surface of the anterior sucker. Here, however, it should be remarked that in most’ cases I have not been able to follow the neck of the gland-cells through the investing membrane of the body to the external surface, but only wp to the basement mem- brane. In T’. ovale, on the other hand, the opening of the glands are distinctly seen at the top of most of the conical papillae already men- tioned. This leads me to suppose that the ducts of the glands through the investing membrane is usually entirely collapsed, being open only during the actual passage of the secretion. It also leads me to suspect whether Brandes” has not mistaken these glands for his “ subcuticular glands.” In two species of Tristomum, viz. T. sinuatum and T. biparasiticum, there is, as before mentioned, a series of groups of peculiar unicellular - glands opening at intervals along the lateral margin of the body. The series begins near the anterior end of the body, and terminates quite near the posterior sucker (Pl. XX, fig. 1; Pl. XXV, fig. 5). In T.. sinuatum I have counted as many as fifty-eight on one side of the body, while in T. biparasiticum there were about sixty-two. I have studied the histology of these glands mainly in the former species, so that the following statements refer mainly to it alone; but the es- sential features are the same in both species, the difference lying only in insignificant details. The cells that constitute these glands are of various sizes, but are more or less polyhedral in form, owing to mutual pressure. ach cell has a distinct wall, and in the specimens I have examined, the greater 1). Brandes—i. c. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 4] part of its interior was occupied by one or more clear vacuoles. The nucleus, which enclosed a single nucleolus, sometimes occupied a central position, but usually lay more or less towards the periphery of the cell ; and in most cases was surrounded by a mass of granular cyto- plasm, which was stained in haematoxylin more deeply than the nucleus itself. From this central mass the cytoplasm radiated, in most cases, in the form of a few threads towards the cell-wall (PI. XXI, figs. 2 & 3). Each cell is prolonged at one of its corners into a long duct ; and the ducts of a single group, after proceeding together for a short distance towards the lateral border, are divided into two bundles, which diverge from each other, and again unite just before opening on the dorsal surface of the body, close to its lateral margin. From the above description, the glandular nature of these cells is, I believe, beyond doubt ; and comparing them with the cells of the submaxillary glands” of the mammalia, there is such a similarity between the two that it is perhaps allowable to conclude that these unicellular glands secrete mucin and help to attach the body of the parasite more securely to its host. The opening of the gland above described is armed with a chitinous piece, which is represented in surface view in fig. 3, Pl. XX, and in section in its natural position in fig. 8, Pl. XXI for T. sinua- tum, and in fig. 5, Pl. XXV, for T. biparasiticum. In T. rotundum similar chitinous pieces (Pl. XXIV, fig. 7) are arranged in transverse rows along the lateral margin of the body just as in Tr. mole”; a single row consisting of four or five pieces in the middle portion of the body, but of only one or two towards the ends. In this species, 1). Landois.—Lehrbuch der Physiologie des Menschen. 6. Aufl. p. 268-270. 2). Parona e Perugia.—Res ligusticae, VIII. Di alcuni trematodi ectoparassiti di pesci marini. Nota preventiva. Annali del Museo Civico di Storia Naturale di Genova. Ser. 2, vol VIL. p. 741. 49 8. GOTO. ad however, there is no gland connected with these chitinous pieces.” These pieces might be supposed to help the parasite in attaching itself to the host, were it not for the fact that they are on the dorsal side of the body, and are therefore presumably of no use in that respect ; but I am not able to suggest any other use of them. | If, now, we compare the structure of the glands hitherto de- scribed, we find a close similarity of their products with one another as well as with those of the sticky glands of Gyrodactylide. The clear, transparent fluid that fills the vacuoles of the lateral glands of T, sinuatum and 1’. biparasiticum. might be mentioned as exceptions ; but we find in their efferent ducts just the same granular substance as in the anterior glands, showing probably that the clear contents of the cells assume the character of granules in their passage along the duct. That the anterior glands of Gyrodactylide are sticky, and are used for attachment can be demonstrated under the microscope; so that I believe it is not much amiss to regard, as I do, the glands above described as having the same function. It may also be mentioned, in addition, that Langley”? and Reid® have described granules in the mucous glands of some ‘vertebrates, which seem to me in many respects similar to those of the glands above described. It is not perhaps out of place here to observe that the position of the opening of the. posterior sticky glands of Monocotyle does not seem 1). Here, again, it should be observed that these so-called chitinons pieces of T. rotundum are ‘deeply stained in borax-carmin. 2). Langley—On the Histology of the Mucous Salivary Glands, and on the Behaviour of their Mucous Constituents. Journ. of Physiology. Vol. X, 1889. p. 433-457. 3). Reid—Mucin Granules of Myxine. Same Journ. Vol. XIV, 1893. p. 340-346. These authors state that the granules are not stained in haematoxylin, and that cold sublimate causes them to burst. This latter fact seems not to accord with the observations described in the text, in which the granules are always preserved intact. Reid also states that boiling water makes the granules break into a mass of stringy débris. Hence we should probably infer that the granules of the glands described in the text and those of the mucous glands of the vertebrates are not exactly of the same chemical constitution. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 43 to accord with their supposed function. A priori, the opening might be expected on the ventral surface of the sucker ; but as the matter stands, I suppose the secretion of the glands flows over the dorsal surface of the sucker on to its margin, and here helps to attach it more firmly to the host. The two genera Heaacotyle and Onchocotyle are destitute of any ‘sticky gland. Hooxs”—These are usually spoken of as of a chitinous nature ; but it should be remembered that they are soluble in a (35 °/,) solution of caustic potash. In the natural state they are slightly yellow, and are very refringent. When they are very slender they are wholly solid ; but when somewhat large they are usually hollow. They lie mostly imbedded in the mesenchyma, and only a very small portion of the pointed end projects free on the surface of the body. ‘Their form, size, number, and position are very varied, but are eminently characteristic of each species. In Onchocotyle, Tristomum, Monocotyle, Calicotyle and Octocotyle major there is only a single pair of hooks. In the last-mentioned species, they are situated at the posterior end of the body between the posterior suckers, close to the median line. They have roughly the form of a fishing hook, and have a process at about the middle of their length, ‘to which is attached a strong bundle of muscular fibres (Pl. IX, figs. 2 & 9 a). In the natural position the pointed end is usually directed posteriorly, but is turned more or less in other directions according to the different states of contraction of the muscle attached to the hook. In QOctoc. major the hooks are hollow towards the pointed end (Pl. IX, fig. 2) ; but in Octoc. minor they are entirely solid (fig. 9 a). 1). For more minute descriptions of the hooks in different species see the systematic portion of the present paper. 44 8. GOTO, In Onchocotyle, the hooks are situated near the extremity of the caudal appendage, between the pair of small suckers already described, on the morphologically ventral side (Pl. XV, figs. 1 and 2). They are wholly solid, and are like a fishing hook in shape ; but the basal end is divided into two processes, one of which bears again a rounded process (Pl. XV, fig. 5). I have not been able to demonstrate any muscular fibres attached to the hooks. In Tristomum, the hooks are situated at the extremities of the posterior border of the central polygon of the posterior sucker, and are provided each with a strong muscle formed by the direct continuations of the longitudinal fibres of the body. They are,unlike the hooks of most other genera, more or less straight (Pls. XX, XXIII, KXLY, XXYV). In Calicotyle and Monocotyle, the hooks are situated in the posterior radial spokes of the posterior sucker on both sides of the median line of the body, and project free over the surface of the invest- ing membrane at the margin of the sucker (PJ. XVII, figs. 1 and 2; Pl. XIX, fig. 1). They are strongly recurved in both genera. In Hezxacotyle and Octoc. minor, there are two pairs of hooks, of which one is much smaller than the other. They are situated at the posterior end of the body, on both sides of the median line, and between the innermost pair of suckers in Hexacotyle. In both cases the smaller pair of hooks lie nearer the median line (PI. IX, fig. 7; Pl. XII, figs. 1, 2, 4, & 5). Finally in Epibdella, there are three pairs of hooks. ‘They are situated on the ventral side of the posterior sucker, and the three pairs are arranged one behind the other ; the most posterior pair being close to the margin of the sucker. Each hook is provided with a strong bundle of muscular fibres. ; The genera Microcotyle, Axine, and Diclidophora are, so far as I haye observed, entirely destitute of hooks. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 45 5. The Mesenchyma. The mesenchyma of the Trematodes has been variously described by different authors. ‘The fact is that it presents very different aspects in different species and even in different parts of the same specimen. The mesenchyma of the monogenetic Trematodes is, generally speak- ing, of very different appearance from the typical form of the same tissue in the Digenea, which consists of large vacuolated cells, between which fibrous network with small nuclei is present; and varies from a truly cellular character to that of the typical, reticulated, fibrous connective tissue on the one hand and a true syncytium on the other. The general mesenchyma of the body has been distinguished by certain writers into two parts, for which different names have been proposed by different writers, but the terms proposed lately by Brandes,” wz. endo- and ectoparenchyma are the simplest and therefore the most convenient. In Axine heterocerca these two portions are very distinctly separated from each other by a thin membrane of compact connective tissue, and are very different in character (PI. VIII, fig. 1). The ectoparenchyma is, in this species, again dis- tinguishable into two layers, that in which the longitudinal fibres lie and that in which the diagonal and the circular- fibres run. The inner layer, ¢.¢., the one in which the longitudinal fibres run, is, in some places, as thick as 20 4“, and its connective tissue consists of very fine, anastomosing fibres, which are but slightly stained in haema- toxylin, and in the knots of which nuclei are here and there present. The general course of the connective tissue fibres of this layer is at 1). Leuckart—Die thierischen Parasiten des Menschen. If, Aufl, I. Bd., 3. Lief, p. 13 et seq. 2). Brandes—l. c., p. 424, 46 S. GOTO. right angles to its limiting membrane, and in some places, the fibres form somewhat strong, compact bundles, which are more deeply stained than the other parts. The outer layer, viz., that in which run the diagonal and the circular muscular fibres, is generally half as thick as the inner layer. It consists of dense, fibrous connective tissue, scantily inter- spersed with nuclei and deeply stained by haematoxylin. This layer is, in most parts of the body, very distinct from the inner layer, from which it seems to be separated by a very thin membrane ; but in some parts, the two layers are not so distinctly separated from each other (Pl. VIII, fig. 3). In Microcotyle reticulata, also, the external portion of the ecto- parenchyma is clearly distinguishable from the inner ; the former con- sisting, in fact, of a dense, diffusely stained, connective tissue, whose fibres run parallel to the investing membrane (PI. III, fig. 4). In all the other species I have hitherto studied, the ecto- and endoparenchyma are not separated from each other so distinctly as in Axine heterocerca; but in most species the two portions present different aspects, the ectoparenchyma consisting generally of a dense, fibrous, connective tissue which is more deeply and more diffusely stained with haematoxylin than the endoparenchyma. There is, however, as above stated, no distinct landmark that separates the two from each other, and even in Axine heterocerca there is no distinct boundary between them towards the two ends of the body. The endoparenchyma presents, in most species, different aspects in different parts of tne body. In Axine heterocerca (Pl. VIII, fig. 1), the endoparenchyma is, in most parts of the body, of a cellular nature. ‘The cells which are of very different sizes in different parts, have usually a distinct membrane and a deeply stained nucleus enclosing mostly a single, but sometimes two or more nucleoli. The cell-body encloses some large vacuoles, and the protoplasm STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 47 radiates, in consequence, in the form of threads from around the nucleus towards the cell-membrane. In some cases, however, there are no vacuoles, and the cell-body consists simply of « slightly stained, granular protoplasm. In still other cases (¢.7., on the right side of the prostate glands around the vas deferens in fig. 1, Pl. VIII), the cell- membrane is incomplete, and the boundaries of adjacent cells are more or less obliterated. In the terminal portions of the body as well as around the ducts of the genital organs, the endoparenchyma consists of truly fibrous connective tissue interspered with small nuclei, which are sometimes surrounded by a scanty mass of granular protoplasm. Around the ducts of the genital organs I have often observed much larger, oval or elongated nuclei surrounded also with a ~ scanty mass of granular protoplasm ; but they seem to have no special function. The meshes of the connective tissue are filled in some parts with slightly stained granules and:in others with a perfectly trans- parent, clear fluid (Pl. VIII, figs. 2, 8, and 4). In most species of Microcotyle, the greater part of the endo- parenchyma is of a fibrous nature; but in some parts it consists of true cells, while in others there is some tendency in it to assume the character of a syncytium. Thus in Microcotyle truncata (Pl. III, fig. 6) and MM. caudata (Pl. III, fig. 9), the endoparenchyma consists in the lateral part of the median portion of the body free from the vitellarium, often of polyhedral cells each usually with a distinct . membrane, a vacuolated cell-body, and a nucleus in the centre, from which the protoplasm radiates in the form of fibres towards the periphery, just.as in some mesenchyma cells of Amine already de- scribed. In JI. caudata the vacuoles are less distinct, and the whole cell is more granular and more deeply stained than in JL. truncata. The nucleus usually encloses a single nucleolus, but sometimes more ; and the boundaries of the adjacent cells are, in many cases, more or 48 8. GOTO. less obliterated, as may be seen in the figures above referred to. In M. reticulata, the endoparenchyma of which consists generally of a typical, reticulated fibrous connective tissue, there are scattered here and there, sometimes in small groups, cells of a roundish or polyhedral outline, with a distinct membrane and a lightly stained, granular protoplasm (PI. III, fig. 4, mes. c.). Besides these, vacuolated cells whose boundaries can be recognized with difficulty are also to be observed. In Diclidophora sessilis, I have observed in the neighbourhood of the ovary, large vacuolated cells with a centrally situated nucleus, from which granular protoplasmic threads radiated towards the cell- membrane (PI. XI, fig. 7)—cells, therefore, which are very similar to those of the mesenchyma of many distomes. In this species there are also polyhedral cells similar to those described in Af. truncata and M. caudata, between the lateral and the median portion of the body (Pl. XI, fig. 5). In M. chiri (PI. I, fig. 5), M. elegans (PI. V, fig. 2), M. sciaenae (PI. VI, fig. 2), Calicotyle Mitsukurit (PI. XIX, fig. 8), Monocotyle Ijimae (P]. XVIII, fig. 2), Onchocotyle spinacis (P1. XV, 10; Pl. XVI, fig. 8), and Tristomum ovale (PI. XXIII, fig. 7),:and_in those portions of the body of M. fusiformis and M. caudata which adjoin the genital opening, the endoparenchyma consists of a reticulated, fibrous connective tissue with large meshes filled with granules which are in some species faintly, but in others well stained (PJ. V, figs. 1, 2, and 3)—the mesenchyma assuming in the latter case more the character of a syncytium. In some species moreover, as in Onchocotyle spinacis, the nuclei are generally surrounded by a scanty mass of granular, well-stained protoplasm. In Hexacotyle acuta, the greater part of the endoparenchyma consists of cells apparently destitute of any membrane, but with well- ‘stained, granular protoplasm, and each with a distinct nucleus which encloses one or a few nucleoli. ‘These cells are usually separated from STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 49 one another by intervening spaces which in sections appear perfectly colourless ; but in many cases the protoplasm of adjacent cells is con- tinuous, and no boundary can be recognised between them (PI. XIV, fig. 2). In Hexacotyle grossa (Pl. XIV, fig. 6) and Tristomum sinuatum (Pls. XXI and. XXII), the endoparenchyma is a true syncytium traversed, however, by scattered connective tissue fibres. The nuclei are very irregularly. distributed ; and in both species they are much smaller than in others. In Hesacotyle grossa I have observed, scattered here and there apparently without order in the general syncytium, cells of irregular forms, -with a homogeneous protoplasm which was well stained by haematoxylin, but differently from the general syncytium (PI. XIV, fig. 6, x), which was stained deep blue, while the cells above mentioned were stained purplish red, just the colour of acetic haematoxylin. I have occasionally observed similarly stained cells in Onchocotyle spinacis. These cells have, so far as I have observed, no relation with the glands hitherto described or to be described hereafter; and the only suggestion that I can give as to their nature is that they are perhaps worn-out: cells undergoing disintegration, and that their colour is due to the presence of some acids generated by decomposition. In Tristomum sinuatum, I have often observed in the peripheral portion of the endoparenchyma cells (already referred to in describing the investing membrane) of a rounded outline and with a granular protoplasm. I do not know what else they are than remnant cells of the original mesenchyma. Around the terminal portions of the genital ducts, the mesenchyma is specially modified in various ways. These will be treated of in con- nection with the genital organs. If, now, we make a general survey of the various forms of endo- parenchyma found in the monogenetic Trematodes, we see that even in the same species it is very different in different parts of the 50. 8. GOTO. body, while individual variations in this respect may be almost imper- eeptibly small. Near either end of the body the endoparenchyma consists, in almost all species, of typical, fibrous, reticulated, connective tissue, interspersed with nuclei (PI. II, fig. 7; Pl. IV, fig. 2; Pl. XII, figs. 4&7; Pl. XV, fig. 8; Pl. XVI, fig. 1). In other parts of the body we may distinguish two different tendencies in the course of differentiation of the original mesenchyma cells. Some of these cells appear to assume more of a vacuolated character, and these furnish the fibres of the connective tissue ; while others come to have more and more a granular protoplasm without any distinct membrane, and to form finally a continuous syncytium by simple obliteration of their boundaries. These two tendencies prevail in different degrees in different parts of the body, and also in different species. An endo- parenchyma of a typical, reticulated, connective tissue we have found in Microcotyle sciaenae,. Calicotyle Mitsukurii, and in the terminal por- tions of the body of many other species ; that of a true syncytium we have found in Tristomum. sinuatum and Hexacotyle grossa ; while in all the other species the syncytial and fibrous characters are variously intermingled. 6. The Digestive System. ° The digestive system consists of the mouth and its cavity, the pharynx, the oesophagus, the intestine, and the glands which are connected with them. The mouth is a fannel-shaped opening situated, in all the species, on the ventral side near the anterior extremity of the body. In Microcotyle, Axine, Onchocotyle, Diclidophora, Octocotyle, and Hexacotyle, it is close to the anterior end of the body ; while in Tristomum, Epib- della, Monocotyle, and Calicotyle it is more distant from it. In most species there is a tolerably ample mouth-cavity ; and in those forms STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 51 which have paired. anterior suckers within the mouth (‘‘ Mundsaug- nipfe” of Braun), it freely communicates with the cavities of the latter. In Microcotyle (PI. LV, figs. 1 and 2), Octocotyle, Diclidophora, Onchocotyle, and Hexacotyle, the mouth-cavity expands more or less at its posterior end and constitutes the prepharynz, into which the anterior end of the pharynx protrudes. In Tristomum and Epibdella, however, the mouth and the anterior end of the pharynx coincide so- that in these there is no mouth-cavity (Pl. XXI, fig. 1; XXIII, fig. 4; XXIV, fig. 12; XXV, fig. 7). The pharynx is more or Jess ellipsoidal or almost spherical, or it may have a constriction on each side, so that in horizontal optic section it appears double, as in many species of Tristomum (Pl. XXIII, fig. 1; Pl. XXIV, fig. 10; Pl. XXV, figs. 1 and 4). In Microco- tyle, Octocatyle, Diclidophora, and Onchocotyle, it is traversed by a small, tubular canal which puts the oesophagus in communication with the cavity of the mouth, and is lined by a structureless membrane of various thickness according to the species. In Monocotyle, the cavity of the pharynx is triangular in cross-section (PI. XVIII, fig. 8) ; while in Tristomum and Epibdella, it is funnel-shaped and spacious. Considering this internal cavity as the axis for the sake of orientation, we may distinguish the ventral and dorsal halves of the pharynx ; and the ventral half thus distinguished is always smaller than the other half—resembling in this respect the oval anterior sucker of Onchocotyle and the distomes (PI. IV, fig. 1; Pl. XV, fig. 3; Pl. XVIII, fig. 4; Pi, XA, fe: Ly Pl AAU, fie. as Pl. XALY, fe 12; Pl RY, fig. 7). In nearly all species, the pharynx is an independent organ entirely separated from the surrounding mesenchymatous tissue by a distinct membrane of a cuticular appearance. In Tristomum, however, this membrane is incomplete at some points at the posterior ends of the ventral and dorsal halves, and thus affords passage to the 59 8. GOTO. efferent ducts of the numerous glandular cells presently to be described. The internal cavity is in some species lined with a com- paratively thick, structureless membrane of varying thickness, staining well with haematoxylin, as in most species of Microcotyle, Diclido- phora (Pl. XI, fig. 1), and Monocotyle (Pl. XVIII, fig. 8); but in others the lining membrane is. exactly similar to that which separates the pharynx from the surrounding mesenchyma. ‘The greater part of the substance of the pharynx consists of mesenchymatous connective tissue essentially similar to that of the other parts of the body and traversed by numerous muscular fibres—the muscular fibres, how- ever, predominating sometimes to an extraordinary degree and form- ing almost the whole, as in Monocotyle (Pl. XVIII, fig. 8) and Microcotyle reticulata (Pl. IV, figs. 2 and 3). The musculature of the pharynx consists typically of three sets of fibres, viz. the internal and external circular fibres and the radial fibres. In most species, the sets of circular fibres are arranged in a single layer directly inside the limiting. membranes of the pharynx ; but in some species they are more than one layer thick, as in Tristomum sinuatum (Pl. XXI, fig. 1) and Diclidophora sessilis (Pl. XI,’ fig. 1). Moreover in all the species of Tristomum I have studied, these two sets of circular fibres are developed in a special degree at’ the anterior end of the pharynx, and constitute a powerful sphincter (PI. X XI, fig, 1; Pl. XXIII, fig. 4; Pl. XXIV, fig. 12; Pl. XXV, fig. 7). The radial fibres are but weakly developed in Microcotyle, Axine, Octo- cotyle, and Onchocotyle ; indeed, in the last-named genus I have not been able to demonstrate their existence with enough certainty. In Tristomum, Monocotyle, Diclidophora, and Microcotyle reticulata, how- ever, they are more strongly developed. Like the dorso-ventral fibres of the body, the radial fibres divide into smaller branches towards their ends, and are inserted into the limiting membranes of the pharynx. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 53 Between these radial muscular fibres there are numerous nuclei, some of which are exactly similar to those of the general mesenchyma of the body ; but there are also others which are much larger and are usually surrounded by a granular protoplasm (PI. XI, figs. 1 and 2). These I believe to be remnants of cells that have produced the muscular fibres. The arrangement of the muscular fibres of the pharynx is in most species as above described ; but there are numerous departures from the rule. For instance, in Diclidophora sessilis there are longitudinal (or meridional) fibres on the ventral side, occupying a short strip along the median line. In Microcotyle reticulata, again, the various sets of mucular fibres and their arrangements are so anomalous that they require separate description. They may be conveniently dis- tinguished into the longitudinal, the circular, and the oblique fibres. The circular fibres must again be divided into two groups, the internal and the external. The internal circular fibres are situated next the lining membrane of the pharynx at a short distance from it (Pl. IV, figs. 2 & 3, int. c. m.). : They are most strongly developed a little behind the middle of the length of the pharynx, and diminish towards either end. They are divided into strong, compact bundles by the radial fibres, which pass between them and are inserted into the lining membrane of the pharynx. ‘The external circular fibres are present along nearly the whole length of the pharynx, but are especially developed in its posterior half, and are there more numerous than the inner fibres, which they resemble in being divided into numerous, com- pact bundles by the radial fibres. In the anterior half of the pharynx they are less developed, and are divided into smaller bundles by the oblique fibres to be presently described. In the lateral part of the ' pharynx the external circular fibres are situated about midway be- tween the inner and the outer limiting membrane of the pharynx ; 54 8. GOTO. but in the dorsal and ventral parts, they are situated directly inside the external limiting membrane (Pl. IV, fig. 2). Some of the ex- ternal circular fibres deviate more or less from their course, and passing between the bundles of the lateral longitudinal fibres, are inserted into the external limiting membrane. The longitudinal fibres are divided into three groups, which may be called according to their relative positions, the internal, the middle, and the external. The internal longitudinal fibres are present only in the posterior third of the pharynx, and are situated between the inner limiting membrane and the inner circular fibres (Pl. IV, fig. 38, d). They do not form any definite bundle, and are inserted by their ends into the inner limiting membrane of the pharynx. The middle longitudinal fibres are arranged in a single layer directly on the outer side of the inner circular fibres, with which they are intimately associated throughout their whole course (PI. IV, figs. 2 and 3, c). ‘The individual fibres of this group are separated from one another by an intervening space, and have a round outline in cross-section. The outer longitudinal fibres are present only in the posterior two-thirds of the pharynx, and are divided into four groups, crescent-shaped in cross-section, and placed symmetrically in the lateral, dorsal, and ventral parts (Pl. IV, figs. 2 and 3, 6 and i’). ‘The lateral groups are situated directly inside the external limiting membrane of the pharynx; but the dorsal and ventral groups are separated from it by the external circular fibres already described. The constituent fibres of these groups are formed into strong, compact bundles, which are separated from one another by the radial fibres and also to some extent by the external circular fibres (fig. 2). They are inserted, posteriorly into the lining membrane at the end of the pharynx and more anteriorly into the external limiting membrane at various points. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 55 There is another peculiarity in the pharynx of M. reticulata, and that is that its anterior end is prolonged into a tube with pointed end and with a thick wall which is differentiated into two layers unstained by borax-carmin, the outer of which is very refractive. Both these layers are homogeneous and without structure. In all the species of Tristomum that I have: studied, there are numerous papillae on the surface of the cavity of the anterior half of the pharynx ; and in some species, as in Tr. ovale (PI. XXIII, fig. 4) and Tr. sinuatum (Pl. XXI, fig. 1), these papillae are present also at the anterior margin. They mark the openings of numerous unicell- ular glands, which have been called ‘“ K6rnerdriisen” or “ Pharyn- gealdriisen”” by Max Braun.” I shall adopt the latter name and call them pharyngeal glands. The glandular cells themselves are situated, so fur as I have observed, not as Braun states, in the wall of the pharynx, but outside it, that is, in the mesenchyma of the median portion of the body behind the pharynx, as may be clearly seen in fig. 5, Pl. XXI, which represents a horizontal section of the region in question in Tr. sinuatum. They are very numerous, and vary in form from a typical goblet to a more or less oval or polygonal shape. The goblet-shaped cells, the comparatively large efferent ducts of which can be generally traced very distinctly to the wall of the pharynx, have vesicular nuclei each with one or more, faintly stained nucleoli ; both the cell-body and the efferent duct are entirely filled with coarse granules which stain but slightly with haematoxylin. On.the other hand, those which are oval or more or less polygonal in form have each a rather small nucleus with a single, deeply staining nucleolus ; and the cell-body is very finely granular and deeply stained (PI. XXI, fig. 5). Both these and the goblet-shaped cells seem to be entirely destitute of a membrane. 1). Braun— Wiirmer ” in Bronn’s Klassen u. Ordnungen. p. 450. 56 8. GOTO. Though I have not actually observed it, a number of these efferent ducts seem to unite into one on entering the wall of the pharynx at its posterior part; and they finally open into its cavity at the top of the papillae already mentioned. ‘That two or more ducts unite together I infer from. two facts, viz. (1) that while only one duct opens at the top of each papilla the total number of these papillae is ‘far less than that of the glandular cells, and (2) that the efferent ducts in -the wall of the pharynx are very much larger than those of the single cells. Sometimes also, a portion of the duct is swollen into a reservoir which may present a vacuolated appearance (PI. XXIV, fig. 12).' Moreover, the granules seem to disappear on entering the ducts ‘of the pharynx, and to. be-agglomerated into a con- tinuous mass; for the contents of these ducts do not present any structure, but consist of a deeply stained substance which has all the appearance of a mucous secretory product. Braun” states that in these glandular pharynges, such as we find in Tristomum, the pharyn- geal cells or those large cells with a finely granular protoplasm seem. to be absent ; but although I have found them very few, they do not seem to be entirely absent (Pl. XXIII, fig. 4). At the same time it should be remarked, however, that I have failed to recognise the terminal cells of the excretory system, which are stated by Braun to be present in the pharynx. The cavity of the pharynx leads into the oesophagus. This is in most species simple and tubular in form ; but in some species, as in Axine heterocerca, Microcotyle reticulata, and Hesxacotyle, it sends out lateral branches on both sides, the ends of which are in M. reticulata all connected together by a longitudinal branch of the intestine (PI. III, fig. 2). In most species of Microcotyle, and in Octocotyle, Hexaco- tyle, and Onchocotyle it is somewhat long; but in all the other species 1). Braun—l. ¢. p. 450. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 57 I have studied, it is very short or almost entirely wanting. Its internal surface presents just the same aspect as that of the intestine, and therefore will be described with that. In Monocotyle, Calicotyle, and Tristomum, (and probably also in Epibdella) numerous unicellular glands open into the oesophagus at its very beginning (Pl. XX, fig. 8: -PL AAL fe ts Ph AX, fe, 125 Pl ANY, fo. 32, XXIII, fig. 4; Pl. XVIII, fig. 5; Pl. XIX, fig. 8), which have been called the salivary glands. In Monocotyle and Calicotyle, they are situated close around the oesophagus, and in the former also a little behind it (Pl. XVIII, fig. 5; Pl. XIX, fig. 8). In Tristomum, how- ever, they are situated quite in the lateral portion of the body on the dorsal side, and are connected with the ocsophagus by long ducts (PI. XX, fig. 8). The cells present somewhat different aspects (PI. XX, fig. 9; Pl XVIII, fig. 5); they stain well, but in some the eell-body is comparatively finely granular, and the oval, vesicular nucleus encloses in its centre a single, well staining nucleolus, while in others the cell-body is more coarsely granular and the large, round, vesicular nucleus usually encloses a few small nucleoli besides the _large one, all of which stain more weakly than in the first class of cells. Cells of the first class also stain more deeply than those of the second. ‘These differences are no doubt due to the different stages of secretory activity ; for, intermediate forms of all degrees are also present. In Tristomum, the oesophagus is separated from the intestine by valve-like projections of the surrounding tissue into the cavity of the alimentary canal in which sphincter muscular fibres are present. (PJ. XAXI, fig. 1; Pl ARV, fie. 12; PL XY, fic. 7; PI. XXIII, fig. 4). In Microcotyle, Axine, Octocotyle, Diclidophora, Hexacotyle, and Onchocotyle, the salivary glands are absent. 5g 8. GOTO. The oesophagus leads into the intestine. This always consists of two branches, which proceed towards the posterior end of the body, where they unite together. in some species, or in others remain separate. They traverse the body midway between the dorsal and the ventral side, and divide it lengthwise into three, roughly equal areas, in the central one of which is placed the greater part of the genital organs, while in the lateral areas only the vitel- larium is generally present. ‘The two divisions of the intestine send out lateral branches on both sides; but those of the inner side are generally very short, while those of the outer are long and bifurcate repeatedly before reaching the lateral borders of the body. In some again, asin Hexacotyle (P]. XIII), there are, besides the two main branches of the intestine above mentioned, two others which are situated close to the lateral borders of the body, and extend backwards through the anterior two-thirds of the body. In this genus as well as in Micr. reticulata, the branches of the intestine form a close net-work not only in the lateral but also in the median portion of the body. The relative arrangements of the main divisions of the intestine as well as of their branches will however be best gathered . from the accompanying plates and the description of species, so that I may here pass them over, and proceed to the consideration of the intestinal epithelium. Two types of intestine. may be distinguished according to the character of the cells that constitute its epithelium. In the first type, the cavity of the intestine is destitute of any continuous epithelium, but is bounded directly by a tunica propria, on the surface of which lie cells without any membrane, which contain in their protoplasm numerous, yellowish, dark-brown, or almost black granules of strong refractive power. In most species, there is a distinct nucleus, and in some, as in Onchocotyle spinacis (P]. XV, figs. 7 and 10), also a well STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 59 staining protoplasm. In the species just named, the cells are cylindri- ca], sometimes attaining the length of about .05 mm., and are in some parts closely crowded. But in most of the other species having the intestine of this type, the cells are separated from one another by wide intervals, and the protoplasm is almost entirely obscured by pig- ment granules. In Microcotyle I have not been able to demonstrate the existence of the nucleus, although there was a clear space in the centre of each cell. In Onchocotyle, again, there are, besides the tall, cylindrical cells already mentioned, smaller cells with a well staining, finely granular protoplasm which is either entirely destitute of pigment granules or contains fewer granules of smaller size. Some of these cells are again very small, and have a scanty mass of protoplasm around the nucleus, which latter remains nearly constant in size in all the forms of cells. Intestines of the type above described are found in Microcotyle, Axine, Octocotyle, Diclidophora, Hexacotyle, and Onchocotyle. The second type of intestine is similar to that of the distomes, and is found in Tristomum, Epibdella, Monocotyle, Calicotyle, and the Gyrodactylidae. In this type, the intestinal cavity is lined by a con- tinuous epithelium consisting of cells which are all similar to one another. ‘The individual cells are either cubical or cylindrical, and have a distinct membrane except on their free borders where it seems to be entirely wanting. ‘The nucleus is always situated close to the tunica propria. In Monocotyle Ijimae (P]. XVIII, figs. 2 and 5), the boundaries of the cells are, in cross-section, not to be clearly seen in every: case ; but in a surface view they can be distinctly recognised (PI. XVII, fig. 11). In this species, moreover, the protoplasm of the cells stains deeply, and contains numerous, deeply staining, granules, which are smaller than and also very different in appearance from the pigment granules of the first type. In Tristomum, however, the 60 8. GOTO. epithelial cells stain very slightly, are generally very clear, and are also very much smaller than in Monocotyle. Having now described the various parts of the digestive system, I shall proceed to note down a few considerations on the physiology of the accessory glands. In the first place, as to the pharyngeal glands. As is well known, there are in Turbellaria numerous unicellular glands in the mesenchyma around, before, and behind the pharynx. These open mainly on the lips of the pharynx, but according to Lang” also on the inner and outer surfaces. They have been called the salivary glands. According to Graff”, two sorts of these glands are to be distinguished in Rhabdocoelida, viz., those that open into the mouth- cavity (Pharyngealtasche), and those that open into the oesophagus or, where such is wanting, between the pharynx and the intestine, the latter sort alone deserving the name of “‘ echte Speicheldriisen.” Those who have had occasion to observe any turbellarian in the living state will have noticed how firmly it sometimes sticks to the side of the vessel in which it is placed, by means of the pharynx and espe- cially its lips, no doubt by virtue of the secretion of the “salivary glands.”’ Although I have not made any convincing observation on this point, I.believe the pharyngeal glands of Tristomum serve the same purpose, and assist in attaching the parasite securely to the host. Again, in Distomum lanceolatum, unicellular glands with granular con- tents open on the anterior border of the body. Their secretion is regarded by Leuckart® as having “eine reizende Einwirkung auf die Gewebstheile des Wirthes.” The numerous unicellular glands that open into the ventral sucker and on both sides of the pharynx of. 1). Lang—Die Polycladen des Golfes von Neapel. pp. 111, 113, 115, 120. 2). Graff—Monographie der Turbellarien. Rhabdocoelida. p. 99. 3). Leuckart—Die menschlichen Parasiten. II Aufl., I Bd. II. Abtheil. p. 366-367. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 61 the Holostomidae are regarded by Brandes” as having the same function. As to the action of the pharyngeal glands in Tristomum I am not able to say anything definite ; at any rate I have not observed any inflammation on the gill of the host, so that I believe they are to be regarded merely as mucous glands. Next, concerning the salivary glands. These are, in my opinion, clearly to be distinguished from the glands of the same name above referred to in the turbellarians. In Rhabdocoelida, however, there are, as already mentioned, two sorts of salivary glands ; and in this case those that open between the pharynx and the intestine are perhaps analogous to or at any rate a forerunner of the salivary glands of the ectoparasitic Trematodes. In Distomum Westermani (Kerbert) and Dist. palliatum (Looss), there are also true salivary glands opening into the oesophagus. As to the function of these glands I am not in a position to make any assertion; but some light will perhaps be thrown on the question after we have considered the characters of the ‘intestinal epithelium. We have already seen that the intestines may be distinguished into two types according to the character of their epithelium. We shall begin with the first type, viz., that in which the epithelium is discontinuous, and in which the cells contain numerous dark-brown or almost black granules. The chief question on which I shall venture to make a few remarks is about the nature of these pigment granules. On this point there are, so far as I know, two views. According to Taschenberg” these pigment granules are the food-particles taken in by these cells from the cavity of the intestine ; and in support of this view he cites his observation that these pigment granules are most 1). Brandes—Die Familie der Holostomiden. Zoolog. Jahrbiicher. Abtheil. f. Syst. Geo. u. Biologie. Bd. 5, 1891. p. 553-561. 2). Taschenberg—Weitere Beitriige zur Kenntniss ectoparasit. mariner Trematoden, 1879. p. 11-12. 62 8. GOTO. abundant when the intestine contains numerous fat-globules, and vice versa, which would not be the case were these granules emptied into the intestinal cavity. The other view regards these granules as zymogenic in nature. This view is probably represented by Zeller”, who says, “ Die Zellen lésen sich mit der Zeit ab und zerfallen. Die abgiingigen werden durch junge ersetzt, welche zwischen jenen sich bilden. Offenbar stehen diese Zellen in ganz bestimmter Beziehung zur Verdauung.”’ A third view is possible, according to which these pigment granules are the indigestible remnants of the food taken in by the cells, whether the food be in the form of granules or whether it be entirely fluid. In my paper on Diplozoon”, I followed Taschen- berg; but further observations and especially a careful comparison of these pigment granules with the granules found in the epithelial cells of Monocotyle have convinced me that the two are not of the same nature. It is, moreover, difficult to understand by what means digestion is carried on in those forms which have the intestine of the first type ; for in these, salivary glands, or any other glands which may be supposed to have a digestive function, are totally wanting. The unicellular glands around the oesophagus described by Zeller in Polystomum integerrimum are, I believe, analogous to the pharyngeal glands of Tristomum ; at least they open at a similar place. Again, in some specimens of Onchocotyle killed with hot sublimate shortly after being detached from the host, I have often observed, in sections, the pigment granules in question in the cavity of the intestine, sometimes in groups and imbedded together in a weakly stained mass. In these cases it is difficult to conceive any external force by which the intes- tinal cells might have been torn away from the tunica propria. It is 1). Zeller—Weiterer Beitrag 4. Kenntn. d. Polystomen. Zeitschr. £. wiss. Zoolog. Bd. 27 1876. p. 241. 2). This Journal, Vol. IV, Pt. I. p. 174. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 63 therefore perhaps allowable to believe that the pigment granules are normally thrown out into the intestine and there furnish the necessary means for the digestion of food. On the other hand, this observation may also be advanced in favour of the third view, according to which the pigment granules would be finally thrown away by way of the mouth. Zeller indeed says that the intestinal cells detach themselves from the wall ; and in this case it is difficult to understand the total absence of any phenomena of division among them. According to my own observation, however, only the pigment granules seem to be thrown out, surrounded by a scanty mass of protoplasm, while the cells themselves in all probability remain collapsed in their former positions, and again resume their activity after the lapse of a certain interval. According to the second view, the smaller cells observed in Onchocotyle with a well staining protoplasm and containing fewer granules of smaller size are to be regarded as those which have not yet arrived at the height of secretory activity ; while according to the third view they are to be regarded as those in which much refuse matter has not yet accumulated. If digestion be regarded as taking place in the cavity of the intestine by means of the pigment granules acting as a ferment, then its product could pass into the tissue only by osmosis and filtration. If, on the other hand, the pigment granules be regarded, according to the third view, as the indigestible remnants of the food taken in by the cells, then digestion has to be regarded as taking place intracell- ularly ; but in this case the essential nature of the process would remain unknown ; and it remains, moreover, to ask in what way the digested food is passed on into the fluid which fills the mesenchyma— a fluid which furnishes in all probability the necessary nutriment to the various organs. I feel myself therefore obliged to leave the nature of the pigment granules undetermined. 64 8. GOTO. In the intestine of the second type, the epithelial cells are, as already stated, all alike, and do not contain pigment granules. In Tristomum and Calicotyle, they stain but slightly, and have seemed in some cases to be more or Jess vacuolated. In Monocotyle Ijimae, on the other hand, they contain numerous, deeply staining granules which are to all intents and purposes exactly similar to the granules found in the cavity of the intestine of the same worm. I therefore believe them to be the products of (partial ?) digestion taken in by the epithelial cells. But by means of what is this (partial ?) digestion carried on ? It has already been mentioned that the cells are all alike in appearance, and none of them present any glandular appearance. In some polyclads (Planoceridae), Lang” has described two forms of cells in the intes- tinal epithelium : those of the first form were elongated and cylindric- al, and usually contained large, homogeneous, refractive granules which stained but slightly and looked like fat-globules; those of the second form were more or less club-shaped, with the thickened end turned towards the lumen of the intestine, and these contained, besides an elongated nucleus, numerous granules which were very deeply stained, and which distinguished themselves from those of the other form of cells by their much smaller size and a regularly spherical shape. ‘The latter cells the above named author regards as secretory cells, and the granules as the product of secretion. In Distomum Westermani, again, Kerbert® found, besides the ordinary intestinal cells, “ kolbenformige Gebilde”’ similar to those of the turbellarians, which however he holds as nothing else than ordinary epithelial cells whose shapes have been changed by the ingestion of food. In Sphy- 1). Lang—i. ec. p. 141. 2). Kerbert—Beitrag zur Kenntniss der Trematoden. Archiv. f. mik. Anatomie. Bd. 19, 1881. p. 552. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 65 ranura, Wright and Macallum” hold that the greater part of the digestive process is effected by the fluid content of the intestine, which they think is of an acid reaction. his they infer from their observa- tion that the swallowed epithelial cells of the host have the form of their nuclear chromatin preserved for a long time, which would not be the case were the intestinal fluid of an alkaline reaction. These writers seem to hold also that the secretion of the digestive enzyme and the ingestion of food eranules—which they concede to take place to some limited extent—are effected by one and the same cells. Now, remembering that even so low in the scale of lifeas in Hydra” the cells of the endoderm are differentiated into two kinds, the secretory and the absorptive, it would seem to many very improbable that these two processes should be carried on in the Trematodes by one and the same cells. Confining our attention to the ectoparasitic Trematodes, we see that in all those forms which have the intestine of the first type, viz., that in which the epithelium is discontinuous, and the cells contain numerous pigment granules, the salivary glands are totally absent ; while in those which have the intestine of the other type they are well developed. Considering this fact in conjunction with the view that the pigment granules of the intestinal cells of the first type are of a zymogenic nature, the thought suggests itself whether the pigment cells of the intestine and the salivary glands are not, in the ectoparasitic Trematodes, vicarious in their functions ; and there seems to be no fact, at least for the time being, that is obnoxious to this assump- tion. In the Planocerida the so-called salivary glands secrete only a sticky fluid, hence the necessity of secretory cells in the intestine. In the Rhabdocoelida, in some of which the salivary glands are, as already 1). Wright and Macallum—Sphyranura Osleri. Journ. of Morphology. Vol. I, 1887 .pp. 33, 35. 2). Cf. Greenwood—On Digestion in Hydra; with Some Observations on the Structure of the Endoderm (Journ. of Physiology, vol. IX, 1888, p. 317—344) and Nussbaum—Ueber die Yeilbarkeit der lebendigen Materie (Arch. f. mik. Anat., Bd. 29, 1887, p. 265-366). 66 8. GOTO, mentioned, divided into two groups, digestion is, according to Graff,” exclusively carried on intracellularly. It is allowable, how- ever, to suspect whether the secretory cells have not been overlooked in some forms, especially in those in which the salivary glands form only a single group opening on the surface of the pharynx. One may, again, regard the “ kolbenformige Gebilde” of Distomum Jester- mani as secretory cells, and bring them forward in opposition to our assumption ; since in this species the true salivary glands are present. But if our assumption be true, there would be a time in the phylogeny of the glands in question when both the true salivary glands and the secretory cells of the intestine would discharge their functions side by side. It may also be said in the way of objection to our assumption, that food granules are observed in the intestinal cells, and that this makes it very probable that these cells have the power of digesting them. But it seems to me that these granules are not the product of a simple disintegration of the swallowed food, but that they have been produced by the action ofa certain (partially ?) digestive fluid ; and moreover it may be suspected whether these granules are not, as sug- gested by Greenwood”, formed secondarily in the cells, and whether they are not taken in primarily in the fluid form. 7. The Excretory System. The excretory system is constructed on the same fundamental plan in all the genera treated of in this paper, and presents only some slight deviations in different genera. In all of them, there are two main vessels on each side of the body, which are directly continuous with each other at the posterior end. They run, roughly speaking, on the ventral side of the intestine between it and the ventral nerve ; 1). Graff—Monographie, p. 96. 2). Greenwood —I. c. p. 323. STUDIES ON THE ECTOPARASITIC TREMATODES OF JAPAN. 67 but as their course is always more or less winding, they are not seen in cross-sections exactly between the intestine and the ventral nerve, but are often situated more on one side, right or left, as the case may be. One of the two main vessels, the larger one, opens to the exterior mostly with an extremely small pore in the anterior part of the body on the dorsal side, close to the Jateral border. This is also the case in Onchocotyle, the paired terminal sacs described by Taschenberg” being, as already stated, nothing else than the suckers. In Microco- tyle, I have not been able to observe the excretory vessels in specimens mounted in toto, and therefore have not drawn them in the figures ; but I could always recognize the two main vessels in serial cross- sections. In this gerfus, the excretory openings are situated nearly on the same level with the genital opening: in WM. elegans it is 5 sections (each=10 x), and in I, caudata 10 sections in front of the genital opening; while in M. sebastis it opens on the same level with it. Lorenz speaks of a small papilla on which the excretory vessel opens ; but I have not observed any in my sections. In this genus, there is no distinct terminal sac, the vessel presenting just a perceptible enlargement before it opens to the exterior. In Diclidophora also, the excretory openings are nearly on the same level with the common genital opening: ec. g., in Diclid. sessilis it is about 7 sections (each = 10 #) behind it ; but in this genus there is a tolerably large, ege- shaped terminal sac (PI. X, fig. 5), which opens directly to the exterior. In