ABSTRACT
Musculo-epithelial cells have been isolated from mesenteries of the sea-anemone Metridium senile, and the descriptions of earlier workers confirmed. The cells contribute both to the muscle-field above the mesogloea and to its overlying epithelium. In sections or whole mounts it is possible to see numerous vertical strands passing from the epithelial elements to their muscle-fibres. The protoplasmic strands are separate from one another and are thus surrounded by fluid which forms a continuous thin layer between the epithelium and muscle-field. It is proposed to call this the subepithelial fluid.
Epithelial elements from contracted mesenteries are much taller than those from stretched tissue. As the area of the mesentery decreases during contraction a reversible change from pavement to columnar epithelium takes place. The epithelium is able to follow rapid contractions without delay, owing to the hydrostatic action of the subepithelial fluid in thrusting it outward. There is as yet no evidence that the epithelial protoplasm moves by its own activity during contraction or relaxation. It may be moved passively and has considerable elasticity. Modifications of the musculo-epithelium in certain anatomical regions are discussed.
Although true musculo-epithelium characterizes only the coelenterates, analogous systems occur in the tissues of several higher animals, and it is suggested that intercellular fluid may have a hydrostatic function in these situations also. The possible metabolic role of subepithelial fluid in Metridium is discussed, and it is suggested that it and the mesogloeal fluid together form an ‘internal medium’ which may provide some degree of biochemical co-ordination in this animal.
Introduction
Mtetridium Senile is an acontiarian sea-anemone (Stephenson, -L VI 1935), whose behaviour and anatomy have received considerable study (Parker, 1919; Parker and Titus, 1916; Batham and Pantin, 1950a-c, 1951, 1954; Pantin, 1952). Its normal activity has a much slower time-scale than that of higher animals, and an inactive appearance may conceal continuous changes in shape which, although extensive, are too slow for the eye to follow. The movements observed in prolonged cycles of activity and in immediate reactions are due to the interplay of various sets of muscles, which contract against the slight pressure of the coelenteric fluid and cause the animal to lengthen or to shorten accordingly. While the nervous and sensory systems play a large part in this activity and must be considered in any physiological interpretation of sea-anemone behaviour, the contractile properties and structure of the muscles are equally important in the living animal. Batham and Pantin’s work in this field has already thrown light on many aspects of the muscular system in Metridium. The cells of this primitive tissue are organized on a basis different from that in higher animals, and as they have not yet been examined from a functional point of view, it is now proposed to do so.
Since Metridium is diploblastic, the muscles develop in single layers over various anatomical surfaces. The fibres run parallel, forming the extremely drawn-out lattice work known as a muscle-field (Batham and Pantin, 1951). This can give rise to more elaborate muscles only by folding, as in the mesenteric retractors, or by rolling up into cylindrical bundles which penetrate the mesogloea, as in the sphincter (fig. 3, c). The muscle-fibres can contract to a quarter or less of their extended length, and appear to be restricted in their excursions by the special properties of the mesogloea immediately beneath (Batham and Pantin). The rest of the mesogloea is strikingly extensible, owing to its fibrous crossed-lattice structure (Chapman, 1953). Above the muscle-field lies an epithelium which has been little studied, although it is able to follow large and occasionally rapid contractions and must possess interesting properties as a plastic covering layer. The organization of the epithelium and its relation to the underlying muscle-field will be considered below.
The Structure Of Musculo-epithelium
As long ago as 1879, the brothers Hertwig showed that muscle-fibres in the endoderm of actinians were really part of epithelial cells, and one of their illustrations is reproduced in fig. 1, A. The cell-body of each muscle fibre was found in the epithelium, and the two layers formed one system. The Hertwigs, interested in the evolutionary aspects of cell-structure, pointed out that this primitive condition could be modified by stages in which the muscle-fibre was developed at the expense of the rest of the cell. In the large muscle-fibres of the sphincter, sunk into the mesogloea (fig. 3, c), the cell-body is reduced to a nucleus and a small mass of cytoplasm. The same is true of the ectodermal muscles of the disk and tentacles in Metridium, and in this case separate epithelial cells provide a covering layer. True musculo-epithelium is present throughout most of the endoderm, however, and the question of immediate interest is its functional activity in these primitive metazoa. It is relevant to consider the fine structure of the cells, their attachment to each other and to the mesogloea, and their behaviour in living tissues during contraction. Later on it may be possible to approach the problem of morphogenesis and to find out how such a system develops, grows, regenerates when damaged, or regresses during starvation. Hydra, for instance, possesses a reservoir of undifferentiated interstitial cells, but there is at present no suggestion of anything comparable in sea-anemones, and little is known of how new tissues arise.
The following account applies to cells from the mesenteries of Metridium, which provide endodermal tissues easy to prepare and to examine (see Batham and Pantin, 1951). The cells described by the Hertwigs can be isolated by their osmic-acetic technique, or by macerating with Goodrich’s fluid (1942); a small one from a stretched mesentery is shown in fig. 1, B. The epithelial part of the cell contains an oval nucleus in granular cytoplasm, and carries one long flagellum at its free surface. The base fans out as a crest along a réfringent muscle-fibre, which is thus wholly invested with a fine layer of cytoplasm. In addition, small cytoplasmic processes may project irregularly beneath the muscle-fibre, which is probably attached by this means to the underlying mesogloea, and perhaps to its neighbours as well (this feature is not shown in fig. 1, B): Semai van Gansen (1952) makes a similar observation for Hydra. At least some of the small granules in the epithelial cytoplasm are mitochondria, for they stain with janus green B in living tissues (Lazarow and Cooperstein, 1953): similar bodies are again present in the endodermal cells of Hydra (Semai van Gansen 1954).
Contracted mesenteries yield musculo-epithelial cells which are very much taller than the one in fig. 1, B, and a typically elongated cell of comparable size is shown in fig. 1, c. In very contracted tissues the epithelial part of the cell is even more attenuated, and the nucleus appears as a bead in a thread of cytoplasm. The portion bearing the flagellum usually remains slightly clubshaped. These changes were described by the Hertwigs and are those which might be expected if the epithelium accommodated to contraction of its underlying muscle-fibres by changes in height. (It should be noted, however, that fig. i, A, reproduced from the Hertwigs, does not illustrate one cell in several phases of contraction, but shows the difference between cells from the crests and troughs of permanent folds in the retractor muscle. A figure of contracted and relaxed musculo-epithelium cells from a tentacle of Sagartia is given in Taf. VI, fig. 11, of the monograph.)
The features characteristic of musculo-epithelial cells can also be seen in whole mounts of Metridium mesenteries, especially after staining with Heiden-hains’ iron haematoxylin (Pantin, 1946). The iron haematein lake is deposited on almost every contour, and suitable differentiation of a whole mount thus reveals both intracellular and surface structures. Baker’s formaldehyde-cal-cium (1944) is a good fixative for these tissues.
Among the most evident characteristics of an intact mesentery is the forest of long flagella which stand out from the surface as shown in fig. i, D. They may be 30 μ, long and appear tangled or straight, depending probably on the speed of fixation. The basal granule of each may be seen just below the epithelial surface, as in fig. 1, E. In life every flagellum beats rapidly, and a current flows over the surface of the mesentery. Most of the flagella appear to lash backwards and forwards in one plane like cilia, but it is interesting to note that those which move slowly (whether naturally or as a result of damage is not clear) sometimes undulate instead. Neighbouring flagella are not co-ordinated, but all beat in the same direction.
It is customary to refer to the ‘ciliary currents’ of actinians, and the term ‘ciliary reversal’ is well known (Parker, 1919). These physiological attributes are, however, largely due to cells which possess only one long ‘cilium’ each. The musculo-epithelial cells are thus, strictly speaking, flagellated, and in this account the conventional word flagellum will be used. It seems difficult to distinguish between cilia and flagella on structural grounds (Bradfield, 1955), and there is probably little difference in their fundamental properties.
The cell boundaries of stretched mesenteries are also revealed by iron haematoxylin. The epithelium consists of polygonal units, each with its flagellum and nucleus, as in fig. 1, F. As has been seen in macerations, the dimensions of the epithelial elements vary with the state of contraction of the mesentery.
The muscle-fibres beneath the epithelium also stain well with haematoxylin, as shown by Batham and Pantin (1951). Individual fibres can be traced in the network of the muscle-field and usually meet with one another at their ends. It is unlikely that they are in cytoplasmic continuity, because when living tissues are treated with methylene blue, the fibres stain singly and not as a syncytium. The muscle-field nevertheless functions as a whole in many of the animal’s activities, and the problem of how separate fibres combine to form a supracel-lular complex is an important one. There is no evidence so far that a true syncytium is present, although contraction often involves the whole tissue. Macerations show clearly the connexion between the epithelium and each fibre, however, as a more or less attenuated thread of protoplasm. This is also evident in whole mesenteries wherever a rent in the epithelium has exposed the muscle-field. The gap is then spanned by numerous protoplasmic stems connecting muscle fibres to their cell-bodies. One such region is shown in fig. i, G, which may be compared with fig. 1, B. These features can also be distinguished in living tissues, and fig. 1, H shows muscle-fibres and their epithelium at the edge of a similar tear in a living, unstained mesentery. It illustrates the striking tensile properties of the musculo-epithelial ‘stems’ when stretched.
Sections of fixed material also show the link between muscle-fibres and the epithelium, as in fig. 1, 1. It is well demonstrated in mesenteries cut slightly obliquely, especially when the epithelium happens to be slightly displaced during fixation.
A diagram attempting to relate the observations made on extended and contracted tissues is given in fig. 2. It is clear that the epithelial elements form a coherent mosaic and, by definition, each element is completely surrounded by its neighbours. The underlying muscle-fibres, however, form a latticework, and here each element runs separately for most of its length. Protoplasmic connexions from the epithelial cell-bodies must therefore diverge as they approach the long, thin muscle-fibres. That is, owing to the geometry of the situation, the protoplasmic stems must lie in a space, and fig. i, G-i, shows that in fact they do so. The musculo-epithelial cells are thus anchored above and below, while their central portions are free and lie in a space. Observation suggests that however small the space in this system may be, it extends continuously between the epithelium and the muscle-field. In living tissues such a space undoubtedly contains fluid. The physical and chemical properties of this subepithelial fluid would confer upon it several functions, of which one would be to act as a hydrostatic layer whenever the muscle contracts. These features do not appear to have been described from higher animals. They are discussed more fully below.
Changes Associated With Contraction
It seems likely that the four-or fivefold changes in length undergone by contracting muscle-fibres in Metridium will involve corresponding reductions in area of the overlying epithelium. It has been seen that the epithelial elements adapt themselves during contraction by becoming tall and thin, so that a reversible change from cubic to columnar epithelium takes place, as suggested in fig. 2. In this system the layer of subepithelial fluid also changes in depth because it is incompressible. As the extended muscle-field contracts, its decreasing area must cause a rise in pressure in the fluid layer : and although this will be transmitted in every direction, its only effect will be to thrust the cell-bodies outward, forcing them to elongate. The epithelium thus follows closely any contraction of the muscle-field, owing to the hydrostatic action of the fluid layer. In principle this mechanism will function whether the volume of fluid present is large or small, but in practice the actual volume will probably affect the speed with which the epithelium accommodates, because such fluid is less viscous than protoplasm. In rapid contractions of the whole tissue it would be an advantage for the volume of subepithelial fluid to be appreciable in relation to that of the epithelial protoplasm (see p. 273). The mechanical properties of the fluid layer would then enable the epithelium to function as a whole without excessive strains. This it appears to do, for instance, above the retractor muscle-fibres.
When fibres of the muscle-field itself are displaced by contraction of a reciprocal muscle, they accommodate by becoming buckled at right angles to the fibre axis (Batham and Pantin, 1951). The circular muscle of the tentacles or body-wall, for instance, becomes buckled whenever the longitudinal muscles contract. If contraction is maximal, or if adjacent muscle-fibres are not in the same state of tone, the epithelial layer also may become buckled. Conducted contractions, for example, may thus be accompanied by fine wrinkling of the epithelial surface, a condition which is often seen in the tentacles, where there is not much subepithelial fluid, and in other tissues such as the mesenteries. While it seems that the epithelium and muscle-field can buckle independently to a certain extent, the dynamics of the process have yet to be studied.
The hydrostatic function of the subepithelial fluid can only be maintained if it is sealed off from the external medium, and in this context the intercellular cement of the epithelium is an important contributing factor. In cases which have been investigated, intercellular cement appears to be part polysaccharide and part protein in nature (e.g., Essner, Sato, and Belkin, 1954), and this is probably true also of actinian material. The cell boundaries are revealed by Harmer’s silver method (1884). They lend themselves to considerable stretching, and it seems likely that rapid epithelial changes are assisted also by elasticity of the cell surface. Micromanipulation with a glass needle certainly shows that fresh epithelium is very extensible and elastic. Also the decrease in total surface area of the mesentery which is associated with contraction could be achieved most rapidly if each epithelial cell bulged outwards as indicated in fig. 2 : this suggestion accords with present observations.
When the muscle-fibres relax the epithelium returns smoothly to its original height. It is not clear whether the cell-bodies assist this process by shortening actively, and they are unfortunately too small to yield any useful information with polarized light. Relaxation is generally helped by the slight coelenteric pressure which acts on the whole animal. It is nearly always slower than contraction, and it is probably affected by viscosity of the protoplasm (and of the mesogloea) rather than by pressure changes in the subepithelial fluid.
It may be concluded from these observations that each musculo-epithelial cell forms part of a major effector system which functions as a whole in the living animal. A relatively small volume of fluid, forming a layer between the epithelium and muscle-field, possesses hydrostatic properties which become important during contraction. It enables the epithelial elements to follow contracting muscle-fibres without delay and ensures that they are not distorted by excessive local strains. It may be noted that sensory, secretory, and other cells found together with musculo-epithelium will undergo parallel changes in shape. There is as yet no evidence that the epithelium plays an active part in re-extension, although the possibility should be borne in mind.
Micro-injection Experiments
The presence of a subepithelial space in normal tissues has been confirmed by micro-injection experiments on living mesenteries. Coloured organic or mineral oil, introduced just beneath the epithelium with a sufficiently fine micro-pipette, does not pass through the muscle-field into the mesogloea, although it can spread quite a long way into the subepithelial space and distend the epithelium considerably. It is interesting that oil can also be injected into the mesogloea without difficulty, and its situation recognized at once. These observations have been made on mesenteries stretched over a glass window in a wax plate. A low-power objective can be used for examination.
Frozen sections of injected mesenteries (fixed in Baker’s formaldehydecalcium) confirm the presence of oil globules between epithelium and muscle, or in the mesogloea as the case may be. This method suggests that the subepithelial fluid occupies a fairly restricted space, even above the retractor muscle. The necessary measurements are difficult to obtain because the tissues vary in thickness, but in frozen sections of fresh or formalin-fixed contracted mesenteries, the subepithelium is itself at least 50 p, deep. These rough figures suggest that the fluid occupies perhaps 4% of the musculo-epithelium by volume. While it is true that sections of specimens prepared by various methods may show a subepithelial space so large that it represents a quarter or even half of the musculo-epithelium (fig. 1,1), this condition appears to be irregular rather than normal, and is probably caused by distortion of the tissues during treatment (compare fig. 1, H). Musculo-epithelium is developed in varying degrees in different parts of the anemone, but it is probably correct to estimate the volume of subepithelial fluid as less than 10% and more than i % of the musculo-epithelium in most actinian tissues. The highest values are probably associated with rapidly conducting muscles such as the retractors and sphincter (see p. 273).
Discussion
Before the organization of musculo-epithelium is discussed, let us consider some probable metabolic functions of the subepithelial fluid. Attention is drawn to the wandering cells which are present, often abundantly, throughout the epithelium and mesogloea, and almost invariably populate the subepithelial space. They are fairly small cells with round nuclei, and the cytoplasm is characteristically filled with fine droplets. The form of these amoebocytes is very variable in different preparations, but surprisingly uniform within one preparation, suggesting that the cells are of one kind only. This is seen in fig. 3, A, showing part of a mesentery stained with chlorazol black E. Amoe-bocytes may sometimes aggregate locally, as illustrated by the section shown in fig. 3, B.
It was shown by Chapman (1953) that amoebocytes are probably not responsible for laying down mesogloeal fibres, although their granulated cytoplasm suggests a secretory function. But they could very well have a role similar to those in Scyphomedusae, which Metchnikoff (1892) found to be concerned with wound-healing and tissue reorganization. They may form a physiological system which extends throughout the body of the sea-anemone. In this case, the intercellular fluid of the tissues, comprising not only the subepithelial layer but also the slightly hypertonic fluid of the mesogloea (see Chapman), could provide a continuous transport medium in which the amoebocytes might function. The passage of materials such as dissolved food and excretory products between endoderm and ectoderm, for example, and the reversible changes in the mesogloea which accompany growth, or regression during starvation, could perhaps be mediated by enzymes from these cells.
It is suggested that subepithelial fluid, while having a special hydrostatic function, is part of the general internal medium of sea-anemone tissues. It was hoped to demonstrate its continuity with the mesogloeal fluid by observing the diffusion of aqueous dyes after micro-injection, but this procedure has not been satisfactory. Nevertheless, since tissue fluid facilitates metabolic exchanges between neighbouring cells, in this case it would certainly promote the local integration of biochemical processes, especially in animals as large as sea-anemones, which have no circulatory system within the tissues. Diffusion could proceed readily in a fluid layer of these dimensions. There is at present insufficient knowledge of metabolic functions in actinians for a detailed discussion of the subject: but it may be pointed out that owing to its distribution the subepithelial fluid provides the immediate environment for the axons of nerve and sensory cells, of whose activities much is already known (Pantin, 1952), and for all the epithelial elements such as mucus cells and nematocysts which are found in anemone tissues together with musculo-epithelium. The cells of the nerve net actually occupy some of the subepithelial space, since they lie above the muscle field and run between the bases of the epithelial elements.
Musculo-epithelial cells occur widely among cnidarian coelenterates. They were first described by Kleinenberg (1872), whose discovery of the ‘neuromuscular’ cells of Hydra was much discussed at the time. They have since been found in medusae (Krasinka, 1914), in Lucernaria (Korotneff, 1876), Veretillum (Bujor, 1901), and Gorgonia (Chester, 1913), and Antipatharia (Dantan, 1920) and other orders of coelenterates.
More recent studies of Hydra (Goodrich, 1942; Mueller, 1950; Semai van Gansen, 1952) have shown that the isolated musculo-epithelial cells resemble on a smaller scale those of Metridium’, they differ in that the endodermal cells are phagocytic, and that in the ectoderm each cell has several musclefibres and lacks flagella. The muscle layer of Hydra has been variously interpreted as a contractile network (e.g. Mueller) and as part of the mesogloea (Holmes, 1950), but only Hadzi (1909), who was primarily interested in the nervous system, has considered the musculo-epithelium as a whole. He describes large fluid-filled vacuoles in the epithelium, which interconnect through a mesh of vertical protoplasmic strands. His figure suggests a system of subepithelial fluid analogous to that in Metridium. Sections of fixed Hydra, on the other hand, usually show much larger spaces within the cells than between them, and when they are compared to Hadzi’s diagram it is clear that the whole problem needs re-investigating.
It seems likely, nevertheless, that in many coelenterates the musculo-epithelial tissues possess in some degree a layer of subepithelial fluid. It provides a mechanism which overcomes the viscosity of epithelial protoplasm, and therefore gains significance in well-developed muscles which contract rapidly, such as the mesenteric retractor of Metridium’, and in fact, it is usually in such tissues that the subepithelial space is most evident (see also fig. 3, c). In partial or localized contractions a certain amount of shearing movement must take place, and the fluid layer, together with the elastic properties of the cells, will allow this to proceed smoothly.
The actinian muscle-field is modified in special anatomical regions, of which the sphincter is an important example in Metridium. The structure of the mesogloeal sphincter of the related anemone Calliactis parasitica Couch is shown in fig. 3, c. It is primarily a specialization of the body-wall circular muscle, in which numerous folds of the muscle-field have sunk into the mesogloea and become pinched off as separate bundles. Each bundle forms a cylinder : the muscle-fibres line the periphery, and often undergo yet further folding. The cell-bodies of the muscle-fibres face the axis of the cylinder and are reduced in size. Their free epithelial surface has been lost in this situation, but each cylinder nevertheless contains the equivalent of a subepithelial space and is filled with a small amount of fluid. This will assist rapid contractions and it therefore retains a hydrostatic function. Nerve-cells supplying the sphincter, which have not yet been detected histologically although the evidence for them is conclusive (Pantin, 1935), may be expected to run in the subepithelial space as they do in other tissues.
The general circular muscle of the body-wall also divides into bundles wherever it passes beneath the attachment of a mesentery (Batham and Pantin, 1951), as illustrated by the model shown in fig. 4. The mesogloea of the mesentery is continuous with that of the body-wall by short strands, which interdigitate with the circular muscle bundles. The mesogloeal strands are marked Xin figs. 3, D and 4. Fibres of the circular muscle, unlike those of the sphincter, are part of a true musculo-epithelium, and all the cell-bodies contribute to the epithelium covering the muscle-field (omitted in fig. 4 for the sake of clarity). The epithelia of the circular and parietal muscles meet along the junction of mesentery and body-wall, and the parietal fibres seem to abut right on to those of the circular muscle (see fig. 4). The circular muscle fibres do not appear to carry cell-bodies with them as they dive under the mesentery, but the subepithelial space continues through the cylinders and may contain strands of protoplasm. Since the nerve-net runs in the subepithelial space, it is possible to see how various regions of the circular and parietal muscles may be co-ordinated. Such co-ordination is an important feature in much of the slow, reciprocating activity characteristic of sea-anemone behaviour (Batham and Pantin, 1954), and further histological study of this region should prove rewarding.
In triploblastic animals, muscular tissue is built up in depth and does not usually arise from epithelium. There are, nevertheless, records of musculo-epithelium from several groups outside the coelenterates. Among Turbellaria, Prorhynchus is one of several primitive though unrelated genera in which it has been reported (Hyman, 1951), but although the epidermal muscles lie very near the surface, sections of P. putealis Haswell do not show signs of any true musculo-epithelium (fig. 5).
After staining with Mallory, sections of a specimen fixed in Bouin’s fluid (kindly provided by Dr. C. F. A. Pantin) show that while several layers of epidermal muscle are present, they are separated from the outer layer of cytoplasm by an envelope of connective tissue. As may be seen from fig. 5, beneath this ‘basement membrane’, the muscles form four strata of alternating circular and longitudinal fibres, the latter being most strongly developed. The structure of the outer layer of cytoplasm is not clear in this specimen, although it appears to be non-nucleate and must therefore be connected to the underlying tissues : it probably represents the peripheral part of sunken epithelial cells, whose nuclei are sandwiched between the two double layers of muscle. The arrangement of these four layers of muscle beneath a membrane of connective tissue is, however, such that none could form part of a true musculo-epithelium. The epidermis is nevertheless remarkably plastic in such animals, and there may be a functional parallel.
Among annelids, while certain cells lining the blood-vessels of polychaetes can contract independently (Hanson, 1951), true musculo-epithelium again seems to be absent. On the other hand, sections of the earthworm (fig. 6) and also of several polychaetes show that an analogous system may be present in the longitudinal muscle of the intestine. Here a flexible columnar epithelium lies just above a single layer of muscle-fibres, resembling the ectoderm of disk and tentacles in Metridium in structure, and perhaps the musculo-epithelium of the endoderm in function. The blood sinuses of polychaetes must incidentally have a hydrostatic effect wherever they run near the periphery of contractile tissue. In fact, intercellular fluid probably has some mechanical importance in lower invertebrates wherever it fills in the surface layers of loosely packed tissue, and a study of several groups from this point of view would be very interesting.
Vertebrates all possess some form of myoepithelium. It has been studied particularly in sweat and mammary glands, and consists of stellate cells which probably form a network (Richardson, 1949) and contract independently, but they do not resemble coelenterate musculo-epithelium at all. The dilator pupillae of the iris, however, which is also ectodermal in origin, retains a primitive structure in the adult (Heerfordt, 1900) which can appear very similar to that of certain tissues in Metridium. The muscle-fibres are smooth, they are attached to a compact cell-body which forms the overlying pigmented epithelium, and as in the retractor of Metridium (Batham and Pantin, 1951), permanent buckling seems to have occurred. It would be striking to find a functional similarity in such diverse situations.
It may be seen that while true musculo-epithelium is almost entirely confined to the cnidarian coelenterates, analogous tissues may occur in higher animals wherever a flexible epithelium covers unstriated muscle. In addition, the actinian muscular system resembles certain smooth muscles of higher animals in many of its physiological properties (Batham and Pantin, 1951). The excitation and propagation of contractions, however, is not fully understood in either case, and it is to be hoped that future work will throw light on such problems. The present account has emphasized some of the differences between the so-called primitive tissues of coelenterates and those of triplo-blastic animals : but it is likely that even special features such as the hydrostatic role of subepithelial fluid, and the supracellular organization of the muscle-field, have functional parallels among a large number of other organisms.
ACKNOWLEDGEMENTS
This research was carried out during the tenure of a D.S.I.R. assistantship under Dr. C. F. A. Pantin, F.R.S., in the Department of Zoology, Cambridge, and at the Marine Biological Station, Plymouth. It is a pleasure to express my sincerest thanks to Dr. Pantin for his never-failing interest in the work and for much stimulating discussion and advice. I also wish to thank the Director and Staff at Plymouth for providing working facilities and a supply of living material.