Nephtys lacks circular body-wall muscles. The chief antagonists of the longitudinal muscles are the dorso-ventral muscles of the intersegmental body-wall. The worm is restrained from widening when either set of muscles contracts by the combined influence of the ligaments, some of the extrinsic parapodial muscles, and possibly, to a limited extent, by the septal muscles. Although the septa are incomplete, they can and do form a barrier to the transmission of coelomic fluid from one segment to the next under certain conditions, particularly during eversion of the proboscis.

Swimming is by undulatory movements of the body but the distal part of the parapodia execute a power-stroke produced chiefly by the contraction of the acicular muscles. It is suspected that the extrinsic parapodial muscles, all of which are inserted in the proximal half of the parapodium, serve to anchor the parapodial wall at the insertion of the acicular muscles and help to provide a rigid point of insertion for them.

Burrowing is a cyclical process involving the violent eversion of the proboscis which makes a cavity in the sand. The worm is prevented from slipping backwards by the grip the widest segments have on the sides of the burrow. The proboscis is retracted and the worm crawls forward into the cavity it has made. The cycle is then repeated.

Nephtys possesses a unique system of elastic ligaments of unusual structure. The anatomy of the system is described. The function of the ligaments appears to be to restrain the body-wall and parapodia from unnecessary and disadvantageous dilatations during changes of body-shape, and to serve as shock-absorbers against the high, transient, fluid pressures in the coelom, which are thought to accompany the impact of the proboscis against the sand when the worm is burrowing. From what is known of its habits, Nephtys is likely to undertake more burrowing than most other polychaetes.

The ligaments of Nephtys have attracted the attention of several authors (Ehlers, 1864-8; Langerhans, 1880; Schack, 1886; Emery, 1887), but only two have made suggestions about their possible function. Langerhans (1880) thought that the ligaments formed an internal skeleton, a view which has never received support. Emery (1887), thinking the ligaments were striated muscles, supposed from their anatomy that they supplemented the extrinsic parapodial, acicular, and chaetal muscles. Hoyle (1957) disposed of the notion that these structures were muscles by demonstrating their lack of electrical activity, and we have recently shown them to be composed of short lengths of a species of collagen, linked by elastic elements (Clark and Clark, 1960). The function of the ligaments is therefore still unexplained, though the anatomy of the ligamentary system suggests that it is likely to have a restraining influence upon changes in the shape of the worm. If this is so, the ligamentary system must be considered in relation to the anatomy and behaviour of the musculature.

Emery’s (1887) account of the anatomy of the ligamentary system is substantially correct, but the accounts of the musculature of Nephtys by various nineteenth-century anatomists (Ehlers, 1864-8; Rohde, 1885; Schack, 1886; Emery, 1887) are all incomplete. They provide insufficient information for a functional analysis of the musculature to be made, and so we have reexamined the anatomy of the segmental musculature in detail before attempting to consider the function of the ligaments. Two species, N. cirrosa Ehlers and N. hombergi Lamarck, have been used for most of the work, but dissections have also been made of N. bucera Ehlers, N. californiensis Hartman, and N. caeca (Fabricius). No interspecific variation in the anatomy of the musculature has been detected, and from our experiments on the first two species it appears that the behaviour of the musculature is also constant.

The division of the muscles under the headings body-wall musculature, extrinsic and intrinsic parapodial musculature, &c., is intended as a functional classification, but in the last resort, some arbitrary distinctions have had to be made.

Body-wall musculature

The familiar organization of the body-wall of annelids into segmented and paired, dorsal and ventral longitudinal muscles, and a ring of circular muscles, is highly modified in Nephtys. The chief peculiarity, as Ehlers (1864·8), Schack (1886), and Emery (1887) noticed, is the absence of circular bodywall muscles.

Dorsal longitudinal muscles (1) (figs. 1-3). The anterior insertion of these muscles is in the prostomium and the top and sides of the first 5 segments. Parts of these muscles encircle the pharynx, accompanying the circumoesophageal connectives, and are inserted in the lateral lips of the mouth. These muscles taper markedly towards the posterior end of the worm. They are inserted into the dorsal and dorso-lateral intersegmental region and end in front of the anal muscle, in the last segment. The majority of fibres in the muscle-blocks cross the intersegmental boundaries without insertions into the body-wall. This is particularly marked in the anterior 30 to 35 segments, where few fibres are inserted dorsally in the intersegmental region, and the dorsal surface of the worm appears smooth. In the more posterior part of the worm a greater number of fibres have dorsal intersegmental insertions and there the segmental boundaries are marked by transverse grooves corresponding to the muscle insertions. In all segments the chief insertions are dorsolateral, at the top of the grooves between the parapodia. The muscle-fibres are attached to a connective-tissue system which subdivides the muscleblocks into a series of rather indefinite longitudinal compartments. The paired muscles meet in the mid-line and form an almost uninterrupted muscle-block across the top of the segments, and are folded around the sides of the gut.

FIG. 1.

Septal and proximal anterior dorso-ventral (3) muscles, viewed from the posterior side.

FIG. 1.

Septal and proximal anterior dorso-ventral (3) muscles, viewed from the posterior side.

Ventral longitudinal muscles (2) (figs. 1-3). The anterior insertion is in segments 1 to 5 along the sides of the folded membrane which joins the lateral lips of the mouth (see Clark, 195601, fig. 1), and the posterior insertion is in the ventral part of the circum-anal, pygidial muscle-ring. There is a substantial insertion of muscle-fibres into each intersegmental groove on the ventral surface of the worm. The two muscles abut against the ventral nervecord medially and extend to the bases of the parapodia at the sides. Each muscle-block is folded into a U-shape.

Proximal dorso-ventral muscles (3, 4) (figs. 1, 2). Emery (1887) suggested that the dorso-ventral muscles and some of the chaetal muscles represented the circular body-wall muscles. This seems a reasonable interpretation, at least of the dorso-ventral muscles. They appear to be a continuation of the septal musculature on to the body-wall on each side of the intersegmental groove.

FIG. 2.

Proximal posterior dorso-ventral (4) and intestinal (5—7) muscles, viewed from the anterior side.

FIG. 2.

Proximal posterior dorso-ventral (4) and intestinal (5—7) muscles, viewed from the anterior side.

Septal musculature

The septum (figs, i, 2, 3, 12) is attached mid-ventrally to the nerve-cord and laterally to the body-wall in the intersegmental groove. The lower edges of it cross the ventral longitudinal muscles and its upper edge passes beneath the gut and up the outer sides of the dorsal longitudinal muscle-blocks. The septum is not attached to either set of muscles, nor to the intestine. Thus the intersegmental septum is incomplete and the coelomic cavity of one segment is in communication with that of the next around the gut and between the septum and the ventral longitudinal muscles. There are no septa anterior to that between segments 34 and 35, because the anterior segments house the large, muscular, eversible pharynx.

The septum consists of a muscular sheet with an anterior and posterior covering of coelomic epithelium. There is no transverse connective-tissue membrane. The muscle-fibres in the dorsal part of the septum run transversely from body-wall to body-wall, but the majority of the fibres run in a fan from the body-wall to the ventral nerve-cord or to a mid-dorsal projection from it which runs two-thirds of the way up the middle of the septum. The free edges of the septum are thickened and its dorsal edge, beneath the gut, is folded over (fig. 12).

The gut suspension

In the anterior, non-septate segments, the gut is not attached to the bodywall, but is able to move freely and permit the eversion of the proboscis. In all the septate segments the gut is attached to the intersegmental part of the lateral body-wall. There is no medial dorsal mesentary, and no dorsal suspensory muscle to support the intestine; instead, the septa form a series of slings on which the gut rests. It is anchored down on them by 3 pairs of muscles in each segment immediately in front of the septum (figs. 2, 3).

FIG. 3.

Septal and intestinal (5-7) muscles. Antero-lateral view of two segments.

FIG. 3.

Septal and intestinal (5-7) muscles. Antero-lateral view of two segments.

Lateral intestinal muscles (5)

These are inserted along the sides of the gut and run beneath the lateral edges of the dorsal longitudinal muscles to the posterior face of the parapodium, where they are inserted distal to the proximal dorso-ventral muscles (3 and 4).

Superior ventro-lateral intestinal muscles (6)

These are inserted on the gut dorsal to the lateral suspensory muscles and their fibres become continuous with the circular muscle-coat of the intestine. They arise in two groups, a major one posterior to the lateral muscle and a very slight, minor one anterior to it. Both branches partly encircle the gut and join one another at the point where they diverge from the gut-wall and run to the parapodial wall. They are inserted distal to the dorso-ventral muscle (4), but many of the fibres in this intestinal muscle run into and become continuous with the dorso-ventral muscles. The superior ventro-lateral intestinal muscle is the best developed of the muscles inserted on the gut.

Inferior ventro-lateral intestinal muscle (7)

This is inserted on the ventral surface of the gut on each side of the sub-intestinal blood-vessel, and runs to the body-wall at the base of the parapodium.

Extrinsic parapodial muscles

Six muscles arise from the mid-line, where they are inserted on the top and sides of the ventral nerve-cord, and run into each parapodium (figs. 4, 5). Of these the distal anterior and posterior muscles (8, 9) are the best developed. All these muscles are inserted in the neuropodium. The muscles inserted in the notopodium arise on the dorsal body-wall at the sides of the dorsal longitudinal muscles.

FIG. 4.

Extrinsic parapodial musculature and the dorso-ventral muscles. Diagrammatic view of one segment viewed from the mid-sagittal plane towards the parapodium. The acicular muscles have been omitted. Anterior end to the right. 3, proximal anterior dorso-ventral muscle; 4, proximal posterior dorso-ventral muscle; 8, distal anterior parapodial muscle; 9, distal posterior parapodial muscle; 10, proximal anterior parapodial muscle; 11, proximal posterior parapodial muscle; 12, posterior diagonal muscle; 13, trans-septal muscle; 14, dorsal parapodial extensor; 15, dorsal parapodial flexor; 18, distal anterior dorso-ventral muscle; 19, distal posterior dorso-ventral muscle.

FIG. 4.

Extrinsic parapodial musculature and the dorso-ventral muscles. Diagrammatic view of one segment viewed from the mid-sagittal plane towards the parapodium. The acicular muscles have been omitted. Anterior end to the right. 3, proximal anterior dorso-ventral muscle; 4, proximal posterior dorso-ventral muscle; 8, distal anterior parapodial muscle; 9, distal posterior parapodial muscle; 10, proximal anterior parapodial muscle; 11, proximal posterior parapodial muscle; 12, posterior diagonal muscle; 13, trans-septal muscle; 14, dorsal parapodial extensor; 15, dorsal parapodial flexor; 18, distal anterior dorso-ventral muscle; 19, distal posterior dorso-ventral muscle.

FIG. 5.

Extrinsic parapodial musculature. Diagrammatic dorsal view. 3, proximal anterior dorso-ventral muscle; 4, proximal posterior dorsoventral muscle; 8, distal anterior parapodial muscle; 9, distal posterior parapodial muscle; io, proximal anterior parapodial muscle; 11, proximal posterior parapodial muscle; 12, posterior diagonal muscle; 13, transseptal muscle; 18, distal anterior dorso-ventral muscle; 19, distal posterior dorso-ventral muscle.

FIG. 5.

Extrinsic parapodial musculature. Diagrammatic dorsal view. 3, proximal anterior dorso-ventral muscle; 4, proximal posterior dorsoventral muscle; 8, distal anterior parapodial muscle; 9, distal posterior parapodial muscle; io, proximal anterior parapodial muscle; 11, proximal posterior parapodial muscle; 12, posterior diagonal muscle; 13, transseptal muscle; 18, distal anterior dorso-ventral muscle; 19, distal posterior dorso-ventral muscle.

Distal anterior parapodial muscle (8)

This muscle arises near the middle of the segment, where it is inserted on the top of the nerve-cord sheath, and runs to the anterior wall of the parapodium at the outer edge of the distal anterior dorso-ventral muscle (18) at the level of the top of the neuropodium.

Distal posterior parapodial muscle (9)

This originates on the side of the nerve-cord sheath, beneath the anterior muscle (8). It runs to the posterior wall of the parapodium, where it is inserted between the proximal (4) and distal (19) dorso-ventral muscles.

Proximal anterior parapodial muscle (10)

This originates on top of the ventral nerve-cord between the distal parapodial muscles (8, 9) and the septum, and runs to the anterior wall of the parapodium, where it is inserted at the inner edge of the dorso-ventral muscle (18).

Proximal posterior parapodial muscle (11)

Originates immediately posterior to the distal anterior muscle (8), on the top of the nerve-cord sheath, and runs to the intersegmental groove.

Posterior diagonal muscle (12)

This originates at the side of the nerve-cord beneath the proximal anterior muscle (10) and runs diagonally beneath the distal anterior (8) and above the distal posterior (9) muscles to an insertion in the posterior wall of the parapodium near the intersegmental groove.

Trans-septal muscle (13)

This arises in the extreme posterior part of the segment from the side of the nerve-cord sheath. It passes under the septum and is inserted on the anterior wall of the parapodium behind.

Dorsal parapodial extensor muscle (14)

(figs. 4, 6, 7). This originates in the dorsal part of the segment at the outer edge of the longitudinal muscle. Part of its insertion extends on to the anterior face of the parapodium. The muscle runs diagonally across the parapodium to an insertion on its posterior face between the proximal (4) and distal (19) dorso-ventral muscles, at the level of the lower edge of the dorsal longitudinal muscles.

FIG. 6.

Intrinsic parapodial musculature. Muscles on the posterior wall of the parapodium. 14, insertion of dorsal parapodial extensor; 19, distal posterior dorso-ventral muscle; other muscles not named.

FIG. 6.

Intrinsic parapodial musculature. Muscles on the posterior wall of the parapodium. 14, insertion of dorsal parapodial extensor; 19, distal posterior dorso-ventral muscle; other muscles not named.

FIG. 7.

Intrinsic parapodial musculature. Muscles on the anterior wall of the parapodium. 14, dorsal parapodial extensor; 15, insertion of dorsal parapodial flexor; 18, distal anterior dorso-ventral muscle; other muscles not named.

FIG. 7.

Intrinsic parapodial musculature. Muscles on the anterior wall of the parapodium. 14, dorsal parapodial extensor; 15, insertion of dorsal parapodial flexor; 18, distal anterior dorso-ventral muscle; other muscles not named.

Dorsalparapodialflexor muscle (15)

(figs. 4, 7, 8). This muscle lies deeper in the parapodium than the dorsal extensor muscle (14). It originates in the dorsal posterior wall of the parapodium and crosses diagon to a large insertion on the anterior face of the parapodium at the dorsal limit of the distal dorso-ventral muscle (18), and on its outer edge. It is considerably larger than the corresponding extensor muscle (14).

FIG. 8.

Intrinsic parapodial musculature. Muscles in the posterior half of the parapodial cavity. There are similar muscles in the anterior half of the parapodium. 15, dorsal parapodial flexor; 19, distal posterior dorsoventral muscle; other muscles not named.

FIG. 8.

Intrinsic parapodial musculature. Muscles in the posterior half of the parapodial cavity. There are similar muscles in the anterior half of the parapodium. 15, dorsal parapodial flexor; 19, distal posterior dorsoventral muscle; other muscles not named.

The intrinsic parapodial muscles

The intrinsic musculature of the parapodium is complicated and there is little to be gained by naming and describing all the muscles shown in figs. 6, 7, and 8, although the muscles have been numbered in the figures to simplify the following discussion of them.

Some of the muscles running to the tips of the parapodial lobes are inserted into the body-wall at the base of the parapodium, but the majority are inserted into 3 clearly defined regions on the anterior and posterior faces of the parapodium. These insertion areas are visible from the exterior as two horizontal grooves, near the upper and lower edges of the parapodium, and a vertical groove joining them (figs. 1, 2). All these regions of muscle insertion are in the proximal part of the parapodium. Muscles from the dorsal and ventral edges of the parapodium (16, 17) provide a support for the horizontal insertion regions. These supporting muscles are better developed on the posterior than the anterior wall of the parapodium because more muscles are inserted there.

The outer margins of the anterior (18) and posterior (19) distal dorsoventral muscles are indicated by the vertical grooves in the parapodial walls. Both originate in the ventral body-wall beside the longitudinal muscle and run to the dorsal horizontal insertion area, though a slip of the anterior dorsoventral muscle (18) continues medial to the large insertion of the flexor muscle (15) to an insertion in the dorsal body-wall.

Muscles run to the tip of the notopodium from the dorsal body-wall on both the anterior and posterior walls of the parapodium (23, 24) and there are corresponding muscles in the neuropodium (25, 26). Muscles also run from the body-wall to the dorsal and ventral parts of the parapodium (27, 28), but these do not extend into the tips of the notopodium or neuropodium. They are inserted into the posterior face of the parapodium and there is an additional dorsal muscle (29) in the middle of the parapodium. Two muscles on the posterior face of the parapodium are inserted in the upper and lower parts of the inter-ramal region (20, 21), and a third, on the anterior wall of the parapodium (22), runs to the middle of the inter-ramus.

The most conspicuous muscles lying in the cavity of the parapodium are a pair running from the proximal part of the neuropodium to the tip of the notopodium (30, 31), and a single muscle from the proximal part of the notopodium to the tip of the neuropodium (32). The former run against the anterior and posterior parapodial walls, while the latter runs between them in the cavity of the parapodium. It fans out and is inserted over the whole of the ventral part of the neuropodium.

The acicular muscles

There is one notopodial and one neuropodial aciculum in each parapodium. The bases of the acicula are close together and project slightly beyond the limits of the parapodium into the general body cavity. Each is provided with independent retractor and protractor muscles originating in the parapodial walls and inserted on a pad of connective tissue and aciculoblast cells at the base of the aciculum (fig. 9).

FIG. 9.

Acicular musculature. Diagrammatic view of three segments viewed from the midsagittal plane towards the parapodium. The extrinsic parapodial muscles have been omitted. Anterior to the right, A, acicular retractor muscles; B, notopodial protractor muscles; c, neuropodial protractor muscles.

FIG. 9.

Acicular musculature. Diagrammatic view of three segments viewed from the midsagittal plane towards the parapodium. The extrinsic parapodial muscles have been omitted. Anterior to the right, A, acicular retractor muscles; B, notopodial protractor muscles; c, neuropodial protractor muscles.

The retractor muscles are thin and delicate. One slender muscle (35) of the notopodial aciculum runs to the base of the intersegmental groove. The other notopodial (33, 34) and the neuropodial (36, 37) acicular retractor muscles radiate from their respective acicula to the parapodial walls between the proximal and distal dorso-ventral muscles.

The protractor muscles move the aciculum, and so the parapodium, backwards and forwards as well as laterally. They are inserted at the outer edge of the distal dorso-ventral muscles, though the dorsal anterior notopodial protractors (40) are inserted into the dorso-ventral muscle and its fibres become continuous with it. While all the protractor muscles are substantial, those originating in the anterior wall of the parapodium are better developed than the posterior muscles, and the anterior neuropodial protractors (43 — 45) are subdivided into 3 separate muscles originating at different levels in the parapodial wall.

The chaetal musculature

The chaetal sacs lie in the tips of the noto- and neuropodia on either side of their respective acicula. The chaetal protractor muscles run in a cone from the sides of the chaetal sac to the parapodial wall (48). The retractors (49) run from the base of the chaetal sac to the base of the aciculum (fig. 10).

FIG. 10.

Protractor (48) and retractor (49) muscles of the chaetal sac.

FIG. 10.

Protractor (48) and retractor (49) muscles of the chaetal sac.

Fig. 11.

Dorsal view of the ligamentary system (semi-diagrammatic). Anterior end to the right. In the left-hand segment all the ligaments are indicated; in the other segments only ligaments inserted on the bodywall are included. The ventral nerve-cord is stippled.

Fig. 11.

Dorsal view of the ligamentary system (semi-diagrammatic). Anterior end to the right. In the left-hand segment all the ligaments are indicated; in the other segments only ligaments inserted on the bodywall are included. The ventral nerve-cord is stippled.

The anatomy of most of the ligamentary system was accurately described by Emery (1887), so that only a brief account of it need be given here.

The ligaments are fiat straps of connective tissue, partly elastic, partly inelastic, that run along the dorsal surface of the nerve-cord to which they are attached at two points in each segment. From these points of attachment other ligaments run to the body-wall and into the parapodia (fig. 12). They are very conspicuous in the living worm and are glistening white with dark chevron-shaped or transverse striations on them, giving them a superficial resemblance to striated muscle, with which they were confused for many years. A detailed account of their structure has already been given (Clark and Clark, 1960).

The ligaments are not attached directly to the nerve-cord sheath, but to a structureless, secreted attachment node which is, in turn, attached to the connective-tissue sheath investing the nerve-cord. The posterior attachment node is almost on the anterior face of the septum and is tilted upwards so as to be nearly vertical (fig. 12). The anterior insertion node is about threequarters the length of the segment from the posterior node and is horizontal. A third important point of insertion of the ligaments is on the body-wall in the intersegmental groove at the level of the top of the neuropodium.

Three pairs of ligaments arise from the posterior attachment node on the nerve-cord. The most dorsal of these runs into the notopodium and is inserted into the posterior face of the acicular sac. The broader of the two ventral ligaments runs to the insertion on the body-wall in the intersegmental groove immediately posterior to it. The finer ligament runs into the neuropodium and is inserted in its posterior wall, where it is joined by a ligament from the anterior node. From this point ligaments run to the neuropodial aciculum and chaetal sac.

Two pairs of ligaments arise from the anterior attachment node on the nerve-cord. The more anterior of these is much the stouter and runs to the insertion on the intersegmental groove anterior to it, while a branch from it diverges, passes beneath the septum, and runs into the parapodium of the next anterior segment, where it is inserted on the posterior parapodial wall. The more posterior ligament arising from the node runs to the intersegmental insertion posterior to it, and is accompanied by the posterior diagonal muscle (12) (figs. 4, 5).

Three ligaments arise from the insertion on the intersegmental groove of the body-wall and run into the parapodium of the segment behind. Two pass into the neuropodium, where they are inserted in the acicular sac and in the wall of the neuropodium near the ventral cirrus. The third ligament runs dorsally and is inserted in the wall of the notopodium near the chaetal sac. The attachment nodes on the body-wall are connected by a lateral longitudinal ligament on each side of the body. This runs over the ventral longitudinal muscle and is as broad as the longitudinal ligament on the nerve-cord.

A further series of ligaments in the mid-sagittal plane provides some support for the septum (fig. 12). The ligaments are inserted directly into the nerve-cord sheath in the middle of the segment, and run to the extension of the nerve-cord sheath in the middle of the septum. A fine ligament connects their point of attachment on the nerve-cord to the anterior attachment node. There is also a very fine ligament that runs from one septum to the next at a more dorsal level, near the free edge of the septum.

FIG. 12.

Ligaments in the mid-sagittal plane. Anterior end to the right. The ligament attachment nodes on the ventral nerve-cord are shown in black.

FIG. 12.

Ligaments in the mid-sagittal plane. Anterior end to the right. The ligament attachment nodes on the ventral nerve-cord are shown in black.

The inner (coelomic) side of the nerve-cord is bounded by a sheath which is an extension of the epidermal basement membrane. The ligaments, by way of their attachment nodes, the septa, and the extrinsic parapodial muscles are all inserted on the sheath. The sheath is strengthened and supported in this region by thickening and by a row of fibres forming a sort of palisade in the mid-sagittal line connecting the nerve-cord sheath with the cuticle (figs. 12; 14, A, B). In transverse section it may appear as if the nerve-cord is bisected longitudinally by a continuous membrane, but in fact there are spaces between adjacent fibres. The modifications of the nerve-cord sheath for the attachment of muscles and ligaments is comparable to those in the ventral nerve-cord of nereids (Defretin, 1949; Smith, 1957).

Each fibre connecting the sheath to the cuticle is the product of a single cell (fig. 13). The cell-bodies are in contact with the cuticle and are 8 to 10 μ wide at the point of attachment, tapering to 3 to 4 μ wide 60 or 80 μ from the cuticle. The cell is slightly sinuous at its base and the nucleus, which is 20 μ long and 3 to 4 μ across, being situated in the basal part of the cell, follows its convolutions. The fibre is secreted at the cell boundary and runs as a solid homogeneous rod towards the nerve-cord sheath. The distal end of the fibre is frayed out and is inserted over an area perhaps 10 to 15 μ in diameter in the nerve-cord sheath. The fibres are highly birefringent and are straight and give the impression of being rather stiff and rigid. They are spaced at intervals of 10 to 20 μ and run more or less parallel to each other. The only wider gap than this between adjacent fibres is at the base of the septum, where the nerve-cord sheath is drawn up into a narrow cone running in the middle of the septum. Emery (1887) described this extension of the sheath as a nerve supplying the septal muscles. While we cannot be assured that there are no nervous elements in this structure, what Emery saw and described was certainly not a nerve, but an extension of the nerve-cord sheath, hollow at its base and containing neuroglial and matrix cells of the nerve-cord.

FIG. 13.

Bases of the cells which secrete the supporting fibres of the ventral nerve-cord.

FIG. 13.

Bases of the cells which secrete the supporting fibres of the ventral nerve-cord.

FIG. 14.

(plate), A, transverse section of the ventral nerve-cord posterior to the sub-oesophageal ganglion, including a supporting fibre of the nerve-cord. Ligaments (dorsally) and extrinsic parapodial muscles (dorso-laterally) are inserted on the nerve-cord sheath. Nephtys cirrosa, picroformol, Holmes’s silver technique. B, mid-sagittal section of the ventral nerve-cord in the middle region of the body, including supporting fibres of the nerve-cord. Other structures may be identified by reference to fig. 12. N. hombergi, Bouin, Heidenhain’s ‘Azan’. c, enlargement of part of a swimming Nephtys. Anterior end to the right. Note the displacement of the posterior chaetae during the power-stroke of the parapodia.

FIG. 14.

(plate), A, transverse section of the ventral nerve-cord posterior to the sub-oesophageal ganglion, including a supporting fibre of the nerve-cord. Ligaments (dorsally) and extrinsic parapodial muscles (dorso-laterally) are inserted on the nerve-cord sheath. Nephtys cirrosa, picroformol, Holmes’s silver technique. B, mid-sagittal section of the ventral nerve-cord in the middle region of the body, including supporting fibres of the nerve-cord. Other structures may be identified by reference to fig. 12. N. hombergi, Bouin, Heidenhain’s ‘Azan’. c, enlargement of part of a swimming Nephtys. Anterior end to the right. Note the displacement of the posterior chaetae during the power-stroke of the parapodia.

The sheath is thickened on the top and sides. The greatest thickening is on the dorsal surface of the sheath, where most of the muscles as well as the ligament attachment nodes are inserted. The mid-sagittal line of the sheath is particularly thick where the fibres connecting it with the cuticle are inserted. Minute filaments of sheath run into the muscle insertions and also into the attachment nodes of the ligaments.

Most worm-like animals change their shape by the antagonistic contractions of circular and longitudinal muscles of the body-wall, acting about the coelomic fluid (or equivalent deformable but incompressible material) which serves as a hydrostatic skeleton (see e.g. Chapman, 1950, 1958). Nephtys lacks a circular muscle coat, although derivatives of the circular muscles may exist in the dorso-ventral muscles and possibly in some of the intrinsic parapodial muscles. In the absence of a complete ring of circular muscles, changes of shape must be accomplished by different means from those described in other worms.

Many active worms, such as earthworms, have intersegmental septa which are complete or nearly so. In Lumbricus the septa are entire, except for a ventral foramen which can be closed by a sphincter muscle, and they serve to divide the animal into a series of water-tight compartments (Newell, 1950). Under these conditions the changes of shape of one segment are almost independent of those of its neighbours. The septa of Nephtys are incomplete, but evidence is presented later to show that under certain conditions coelomic fluid does not pass from one segment to the next during muscular contractions. The behaviour of the septa cannot be observed directly, but the means by which they become effective barriers to the transmission of fluid pressures or of coelomic fluid from one segment to the next, can be inferred from their anatomy. The ventral, free edge of the septum is closely applied to the ventral longitudinal muscles (the space between the septum and the muscle has been exaggerated in figs. 1 and 2), and the outer edge of the septum is attached to the body-wall at the base of the inter-segmental groove, so that when the septal muscles contract the lower margin of the septum must be drawn tightly against the ventral muscle-blocks. Although it is not attached to the sides of the dorsal longitudinal muscles, there is no space between them. When the transverse septal muscles of the upper free edge of the septum contract, the intestine is raised and pressed tightly into the space between the two dorsal longitudinal muscles, and the cross-sectional shape of the gut is always such as to fill this gap completely. The only remaining space by which one segment is in communication with the next is at the lower lateral edge of the intestine, between it and the lower edges of the dorsal longitudinal muscle. From the anatomy of the intestine, the longitudinal muscles and the septum, it is not clear if it would remain open or not when the septal muscles were contracted. The gut may be slightly dilated by the pressure of the septum against it and fill this gap, or the gut suspensory muscles immediately anterior to the septum may block this space and prevent fluid passing from one segment to the next posterior one, though in this case they would clearly act as a one-way valve. It should be noted that a ligament supports the septum on its anterior face and therefore tends to press it against these muscles. It is also possible that a contraction of these muscles would draw the edges of the intestine out so that they came into contact with the longitudinal muscles.

A single segment, assumed to be hydrostatically isolated from its neighbours by the contracted septa, may be regarded as a rectangular box with muscles acting in the 6 bounding planes. Those in the horizontal plane are the dorsal and ventral longitudinal muscles, those in the vertical plane are the dorsoventral muscles, and those in the transverse plane are the transverse septal muscles, the proximal parapodial (10, 11) and the trans-septal (13) muscles. Theoretically, contraction of the longitudinal muscles stretches the vertical and transverse muscles, and contraction of either or both of the latter restores the segment to its original length.

In fact, measurements from photographic enlargements of narcotized (fully extended) worms, and of the same worms fixed in Bouin’s fluid after recovery from narcotization (contracted), show that there is little change in the width of the segments; changes in length are compensated by changes in height (table 1). This is to be expected. The transverse muscle-fibres of the septa are slight and appear too delicate to cause major changes in the shape of the segment. Further, on theoretical grounds, it is generally supposed that the septal muscles keep the septum taut during changes in the shape of the segment, but play no significant part in producing that change of shape (Newell, 1950). The proximal and trans-septal muscles are more substantial and act in the transverse plane. They are so disposed as to be able to resist lateral forces on the lower half of the segment walls, or to restore the shape of the segmental wall if it is deformed by contractions of muscles in other parts of the segment. There are no transverse muscles in the upper half of the segment, so that recovery from deformation of that part of the segment would be difficult. But this part of the segment is not deformed by contractions of the longitudinal and dorso-ventral muscles, presumably because its shape is determined by the underlying longitudinal muscles, and these are so massive as to be unaffected by changes in the internal fluid pressure except in a longitudinal direction.

TABLE 1.

Average percentage change in the dimensions of the segments on contraction the worm

Average percentage change in the dimensions of the segments on contraction the worm
Average percentage change in the dimensions of the segments on contraction the worm

Thus the dorso-ventral muscles are the chief antagonists of the longitudinal muscles, and contractions in the longitudinal plane of the segment are compensated by an increase in the height of the segment, and conversely. The septal muscles play, at most, a very minor role, and the proximal and transeptal muscles serve to restrict changes in the transverse dimensions of the segment. In this they are aided by the ligaments running from the insertions on the nerve-cord to the insertions on the intersegmental body-wall.

Nephtys is a rapid and efficient swimmer and, when submerged, uses this method of progression instead of crawling over the substratum. The mechanism of swimming is essentially the same as that described in Nereis by Gray (1939). The body is thrown into sinusoidal waves which pass forwards along the body. In Nereis nearly all the segments are involved in this undulatory motion, and the prostomium traces out a similar path to that traced out by all the segments, but in Nephtys the long, muscular pharynx imparts a stiffness to segments 15 to 30, and they and the more anterior segments are not involved in the sinusoidal motion; instead, they execute a lateral pendulous movement (fig. 15, B).

FIG. 15.

(plate), A, Nephtys swimming slowly. B, Nephtys swimming fast (wavelength of undulatory waves reduced, amplitude increased, greater relative movement of the parapodia). c, Nereis swimming slowly.

FIG. 15.

(plate), A, Nephtys swimming slowly. B, Nephtys swimming fast (wavelength of undulatory waves reduced, amplitude increased, greater relative movement of the parapodia). c, Nereis swimming slowly.

FIG. 16.

Diagram showing the inclination of the parapodia of Nereis pelagica in relation to the position of the segments in the locomotory cycle. The undulations pass from left to right.

FIG. 16.

Diagram showing the inclination of the parapodia of Nereis pelagica in relation to the position of the segments in the locomotory cycle. The undulations pass from left to right.

FIG. 17.

Diagram showing the inclination of the parapodia of Nephtys hombergi in relation to the position of the segments in the locomotory cycle. The undulations pass from left to right.

FIG. 17.

Diagram showing the inclination of the parapodia of Nephtys hombergi in relation to the position of the segments in the locomotory cycle. The undulations pass from left to right.

Nearly all long, narrow animals swim in a similar manner. In smoothbodied animals, such as eels, snakes, and leeches, the undulatory waves pass along the body in the opposite direction to the direction of motion of the animal, but in nereidiform polychaetes, the undulatory waves pass in the same direction as that of motion of the whole worm, i.e. from tail to head. Taylor (1952), in his theoretical analysis of the swimming of long and narrow animals, demonstrated that this reversal of the direction of motion of the locomotory waves relative to the direction of motion of the animal, could be related to the roughness of the body-wall. Roughness, such as that imparted by the presence of a series of rigid, lateral lamellae fixed at intervals along the body, is sufficient to cause the reversal of the locomotory waves, so that they pass from tail to head when the animal swims forward. This is the situation in the nereidiform polychaetes.

Gray (1939), in his earlier analysis of cinematographic records of swimming Nereis, had shown that the parapodia move backwards relative to the ground at the moment when the underlying longitudinal muscles are maximally elongated, i.e. when the crest of the undulatory wave reaches the parapodium. In addition to this passive role of the parapodia, resulting from the behaviour of the underlying longitudinal muscles during the passage of the locomotory wave along them, the parapodia also exert an additional back-thrust by executing a backwardly directed power-stroke at the moment when they are moving backwards relative to the substratum. The power-stroke executed by the parapodium is not essential for forward progression, as appears particularly from Taylor’s analysis, and, indeed, Gray suggested that in Nephtys the parapodia were entirely passive. This, if true, would be puzzling, for Nephtys is a much better swimmer than Nereis, which makes very slow progress through the water. In fact, the power-stroke of the parapodia of the former is much more conspicuous than that in Nereis.

A direct comparison of the part played by the parapodia of the two worms can be made by considering the angular movement of the parapodium relative to the segment during the passage of a complete locomotory wave. Comparable conditions in the two worms can be assured by considering animals that are swimming at such a rate that the body is thrown into the same number of locomotory waves and by measuring the movement of the parar podia over a single locomotory wave of the same wavelength: amplitude ratio in each, as in figs. 15, A, c.

In Nereis (figs. 15, c; 16, 18) the parapodia are perpendicular to the surface of the segment when the underlying longitudinal muscles are contracted. As these muscles are relaxed, the parapodium is drawn back in a preparatory stroke and then executes the power-stroke, which begins slowly about halfway up the leading edge of the locomotory wave, then increases in speed and continues until the parapodium is half-way down the trailing edge of the wave. A recovery stroke then restores the parapodium to the perpendicular position. An important fact to notice is that the power-stroke continues over about half the complete locomotory cycle of the segment.

In Nephtys (figs. 15, A; 17, 19), under comparable conditions, the resting position of the parapodium is inclined forward at an angle of 25° to 30°. About half-way up the leading edge of the locomotory wave, the parapodium is drawn further forwards in preparation for the power-stroke, which begins as the parapodium is carried up to the crest of the wave and is completed as soon as the parapodium has passed the crest on to the trailing edge. There is then a slow recovery to the resting position. In Nephtys the power-stroke is executed only while the parapodium is on the leading edge and the crest of the locomotory wave, and it occupies a smaller fraction of the complete cycle than it does in Nereis. This is indicated by the steeper slope of the powerstroke in fig. 19 than in fig. 18.

FIG. 18.

Nereis. Diagram showing the change in inclination of the parapodium to the transverse axis of the segment during its passage through one complete locomotory cycle. Wavelength: amplitude ratio, 3 · 6. Horizontal axis: inclination of the longitudinal axis of the segment to the direction of motion of the worm. Vertical axis: inclination of the parapodium to the transverse axis of the segment. This diagram should be compared with fig. 15, c, from which it is derived. See text for further explanation.

FIG. 18.

Nereis. Diagram showing the change in inclination of the parapodium to the transverse axis of the segment during its passage through one complete locomotory cycle. Wavelength: amplitude ratio, 3 · 6. Horizontal axis: inclination of the longitudinal axis of the segment to the direction of motion of the worm. Vertical axis: inclination of the parapodium to the transverse axis of the segment. This diagram should be compared with fig. 15, c, from which it is derived. See text for further explanation.

FIG. 19.

Nephtys. Diagram showing the change in inclination of the parapodium to the transverse axis of the segment during its passage through one complete locomotory cycle. Wavelength: amplitude ratio, 3 · 5 (i.e. a comparable body form to that of Nereis in fig. 18). This diagram should be compared with fig. 15, A, from which it is derived. See fig. 18 and text for further explanation.

FIG. 19.

Nephtys. Diagram showing the change in inclination of the parapodium to the transverse axis of the segment during its passage through one complete locomotory cycle. Wavelength: amplitude ratio, 3 · 5 (i.e. a comparable body form to that of Nereis in fig. 18). This diagram should be compared with fig. 15, A, from which it is derived. See fig. 18 and text for further explanation.

Faster progression through the water is achieved by increasing the number of locomotory waves into which the body is thrown (fig. 15, B), with the effect that the wavelength: amplitude ratio is decreased. Under these conditions in Nephtys the movement of the parapodium relative to the segment is modified (fig. 20). The resting phase of the parapodial cycle is almost abolished and the recovery stroke tends to merge into the preparatory stroke. Both the angle through which the parapodium is moved and its rate of movement are increased.

FIG. 20.

Nephtys. Diagram showing the change in inclination of the parapodium to the transverse axis of the segment during its passage through one complete locomotory cycle. Wavelength: amplitude ratio, 2 · 7 (i.e. rapid swimming). This diagram should be compared with fig. 15, B; from which it is derived. See fig. 18 and text for further explanation.

FIG. 20.

Nephtys. Diagram showing the change in inclination of the parapodium to the transverse axis of the segment during its passage through one complete locomotory cycle. Wavelength: amplitude ratio, 2 · 7 (i.e. rapid swimming). This diagram should be compared with fig. 15, B; from which it is derived. See fig. 18 and text for further explanation.

Evidence that the movement of the parapodium relative to the segment is an active one and is not passively produced, for instance by local eddy currents produced by the locomotory waves, can be gained from fig. 14, c. The long and rather flexible posterior chaetae of the parapodium become swept in an anterior direction as the parapodium is moved backwards during the powerstroke, whereas the stiffer, anterior chaetae are not. There is thus a relative movement between the tip of the parapodium and the surrounding water which can only be produced by muscular activity.

It will be noticed that the movement of the parapodium in Nephtys is confined to the distal half of the parapodium, while the basal part moves hardly at all in relation to the segment (figs. 14, c; 15, A, B). This can only be due to contractions of the acicular muscles, for none of the extrinsic parapodial muscles is inserted in the distal part of the parapodium (fig. 5). In any case, the extrinsic muscles are poorly disposed for producing powerful movements of the parapodium, for those inserted on its anterior face arise in the anterior part of the segment and those on its posterior face in the posterior part of the segment. Only 3 muscles run diagonally across the segment. One, the slender posterior diagonal muscle (12), is inserted in the intersegmental groove and can have no effect on the movement of the parapodium. The other two are the dorsal flexor and extensor muscles (14, 15), and these are inserted in the basal part of the parapodium. The chief propulsive force is therefore generated by the acicular muscles in conjunction with the longitudinal muscles.

This is not to say that the extrinsic parapodial muscles play no part in swimming. The contraction of the longitudinal muscles during the passage of locomotory waves reduces the volume of one-half of the segment and there is a compensating increase in the volume of the opposite half of it. Compared with the segmental body-wall, the parapodial walls are thin and would be deformed by these pressure changes but for restraining influences. The basal part of the parapodium is restrained by the extrinsic muscles, and the distal part, which does become distended when the longitudinal muscles on the opposite side of the body are contracted, are prevented from too great a dilatation by the ligaments. Since the parapodia are inflated to only a very limited extent, the reduction in volume of one half of the segment by the contraction of the longitudinal muscles is compensated by an increase in the length of the longitudinal muscles on the opposite side of the body.

One of the chief problems in soft-bodied animals is the provision of rigid insertions for the muscles that move such skeletal elements as the animals possess. Without rigid insertions, much of the mechanical advantage gained by the use of skeletal rods is lost. In the case of the acicula of Nephtys, rigid insertions must be provided on the parapodial walls for the acicular muscles to produce the power-stroke of the parapodium. The power-stroke is performed while the underlying longitudinal muscles are being stretched and the parapodium is turgid. The acicular muscles are inserted in the parapodial walls near the insertions of the extrinsic muscles and these points appear to be held taut by the balanced and opposed forces of inflation by hydrostatic pressure and restraint by the extrinsic muscles.

While the same general principles obtain in Nereis, both the arrangement and the performance of the muscles differ from those in Nephtys. In Nereis, as in Nephtys, muscles arising from the ventral nerve-cord and the dorsal body-wall are inserted in the parapodial walls, but in the former worm some of those inserted in the anterior wall of the parapodium arise in the posterior part of the segment, and those on the posterior wall, in the anterior part of the segment (Snodgrass, 1938; Defretin, 1949), and so give a greater mechanical advantage. Further, the ventral muscles inserted in the anterior face of the parapodium run to the tip of the neuropodium (personal observation) and are chiefly responsible for producing the power-stroke. In Nephtys, the comparable muscles are inserted in the basal part of the parapodium and, as we have seen, do not serve the same function. In Nereis the power-stroke involves the whole parapodium, and it is probably effected chiefly by the extrinsic parapodial muscles. It is significant that in the heteronereis, which is a much faster swimmer than the immature worm (Lillie and Just, 1913, &c.), the muscles chiefly responsible for producing the power-stroke are much enlarged (Defretin, 1949; Durchon, 1952). The acicular muscles may play a subsidiary role and augment the effect of the extrinsic muscles, because those inserted on the anterior wall of the parapodium, and so capable of moving the parapodium backwards during a power-stroke, are larger than the posterior muscles.

Under suitable conditions, Nephtys buries itself in the sand very efficiently, and burrows rapidly through it. The initial penetration of the substratum and the progress through it are two distinct and clearly defined processes involving entirely different types of muscular activity.

Penetration of the substratum

When covered by water and provided with a substratum, Nephtys burrows into it by one of two methods. It may perform undulatory swimming movements while lying on the substratum. This does not result in any appreciable forward movement, but it agitates the sand and sweeps it on to the dorsal surface of the worm. Within a short time it is buried and then its behaviour changes and it begins to burrow through the substratum. Alternatively, it may swim through the water and approach the substratum obliquely. When the prostomium is in contact with the sand, the worm executes fast swimming movements which impart a rapid, low-amplitude, lateral movement to the prostomium and anterior segments. The agitation of the sand, combined with the forward thrust due to the swimming motion, results in the prostomium and the first 15 or so segments being buried within a short time. Swimming then ceases and burrowing begins. According to Chapman and Newell (1947), penetration into the substratum in this manner depends upon the exploitation of the thixotropic properties of the sand, which is rendered more easily penetrated by agitation.

It has been demonstrated in another context (Clark, 1956b) that contact between the dorsal surface of the worm and a solid body inhibits swimming activity and both these methods of gaining entry into the sand have in common the fact that they begin with swimming, which ceases when some part of the body is covered with sand, and the locomotory behaviour then changes to burrowing.

Burrowing through the substratum

Burrowing in the sand involves the use of the proboscis. Briefly, the proboscis is suddenly everted, punching a hole in the sand, into which the worm crawls when the proboscis has been retracted. Obviously, this method of burrowing depends upon the worm being anchored when the proboscis is everted; otherwise its impact against the sand would merely push the worm backwards instead of making a hole in the substratum. Accordingly, this type of locomotion is not observed unless at least 10 segments are buried and these provide the necessary anchor. Occasionally a worm placed on damp sand, but not under water, will attempt to penetrate into the sand by everting its proboscis. Unless the sand is extremely damp and quite soft, this is an unsuccessful manœuvre and the worm merely succeeds in raising its anterior segments from the surface of the sand. A similar phenomenon can also be observed when a worm attempts to burrow in coarse sand in an aquarium when the sand grains lack cohesion and the animal is unable to achieve the necessary anchorage.

The eversion of the proboscis is accomplished chiefly by fluid pressure generated by contraction of the longitudinal muscles, assisted, perhaps, by the contraction of the dorso-ventral muscles. The dorsal longitudinal muscles of the anterior 30 to 35 segments are particularly important. The limit of strong contraction of these muscles is indicated by a deep transverse furrow that appears across the dorsal surface in the region of the 30th to 35th segment, where they are inserted. It will be recalled that there are only slight intersegmental insertions of the dorsal longitudinal muscles in the more anterior segments. The first 34 segments which house the inverted proboscis are non-septate, with the result that increases in the fluid pressure are readily transmitted to the anterior end of the worm. The contraction of the dorsal longitudinal muscles draws the prostomium and the dorsal part of the first 5 segments upwards at a sharp angle to the rest of the body. Parts of these muscles are inserted at the sides of these segments and also run around to insertions in the lateral lips of the mouth. When they contract, the lips and the walls of the segments are drawn aside, so permitting the eversion of the large, muscular pharynx through these very much smaller segments. The folded membrane which joins the lateral lips is stretched tight beneath the everted proboscis. The modifications of the anterior nervous and bloodvascular systems in the first 5 segments, where the eversion of the proboscis entails considerable movement and distortion, have been described elsewhere (Clark, 1956a, 1958).

When the proboscis is everted, the force with which it is driven against the substratum is limited by the frictional forces between the worm and the substratum that prevent it slipping backwards. Only the widest segments, from about segment 15 to segments 45-50, are broad enough to reach the sides of the burrow that has been excavated by the proboscis in previous cycles of burrowing activity, and these form an anchor. The parapodia of all the segments are directed backwards when the proboscis is everted. Those of the most posterior segments are pressed tightly against the sides of the body; those of the middle, anchoring segments are held out at a slight angle. Segments 15 to 34 become reduced in volume by the contraction of the longitudinal and dorso-ventral muscles, which drives the coelomic fluid forwards into the everting proboscis, but they do not become appreciably narrower.

The passage of coelomic fluid across the segmental boundaries can be followed under favourable conditions by injecting a strong methylene-blue solution in sea-water into some of the segments. It can then be seen through the body-wall on the ventral side of the parapodia, particularly in nearly mature worms in which the body-wall is partly histolysed and is thinner than in the immature worms. Methylene-blue solution injected by micrometer syringe into a segment frequently passes immediately to the rest of the body, but occasionally it remains localized and observations can then be made on the movement of coelomic fluid when the proboscis is everted. If the worm everts its proboscis after being injected in the anterior segments, the methylene blue does not reach segments posterior to the 35th until after several eversions or after wriggling by the worm. Similarly, methylene blue injected into the middle segments of the body does not appear in the anterior segments when the proboscis is everted. Chapman (1951) has found that there is no increase in the hydrostatic pressure in segments in the middle region of the body during eversion of the proboscis, but that occasionally increases can be detected in the anterior segments. Examples of such increases of pressure are from 8 to 13 cm, from 21 to 24 cm, and from 5 to 15 cm sea-water (Chapman, 1951). Evidently the first few septa are able to isolate the proboscis region from the rest of the body. That the proboscis depends upon the increase in hydrostatic pressure in the anterior coelom can be demonstrated by inserting an open cannula in the coelom of the proboscidial region. This prevents the worm from everting its proboscis, and coelomic fluid is driven through the cannula when it attempts to do so.

Inversion of the proboscis is accomplished by the contractions of the proboscis retractor muscles and the relaxation of the longitudinal muscles. The worm elongates and the segments return to their normal resting volume and shape. Segments 35 to 45 become less dilated, but a hold is still maintained on the sides of the burrow by segments 15 to 35, which are now dilated by the presence of the pharynx within them.

The final stage in the burrowing cycle is movement forwards into the cavity made in the sand by the proboscis. Nephtys appears to be unable to ambulate merely by movements of the parapodia, presumably because the extrinsic parapodial muscles are not well situated for this task and the acicular muscles are not provided with a sufficiently rigid insertion on the parapodial wall unless the body executes an undulatory movement. The anterior 15 segments are thrown into waves, exactly like the locomotory waves of the swimming worm. One result of this is that the parapodia of these narrow segments make contact with the sides of the burrow as they perform their power-stroke when the crest of the undulatory wave reaches them. The wider, successive segments become involved in this locomotory movement also, but segments posterior to the 50th remain passive throughout the burrowing cycle and are dragged forwards. When the prostomium reaches the end of the hole previously made by the proboscis, the entire burrowing cycle is repeated, starting again with the sudden eversion of the proboscis.

The method of burrowing employed by Nephtys differs from that of such worms as Lumbricus and Arenicola in which the mechanics of burrowing have been studied by Gray and Lissmann (1938), Chapman (1950), and Newell (1950), and by Wells (1944, 1948, 1954) and Chapman and Newell (1947), respectively. The most significant differences spring from the fact that the body of Nephtys tapers towards each end, does not have a circular crosssection, and contains no circular muscles. The worm can therefore anchor itself to the substratum by only a comparatively short region of the body and cannot execute peristaltic burrowing movements as both Arenicola and Lumbricus can. Instead of burrowing in a relatively steady and continuous way, as the earthworm does, Nephtys burrows in a series of convulsive jerks. A further advantage enjoyed by Lumbricus but denied to Nephtys is that since the advancing tip of the former is extended by the contraction of the circular muscles which reduce its diameter, the pressure it exerts against the substratum is magnified by being applied over a small area (Newell, 1950). Nephtys is a carnivore and has a large, muscular proboscis, so that the pressure it exerts against the substratum is applied over a wide area. Arenicola is in some respects intermediate between Lumbricus and Nephtys in its method of burrowing. Although it moves by peristaltic contractions of the body-wall, the proboscis is used, either to agitate and soften the sand into which the anterior segments are then thrust (Chapman and Newell, 1947), or to scrape away the sand in front of the advancing tip of the worm (Wells, 1948).

Unfortunately it has not proved possible to investigate experimentally the function of the ligaments while the worm is burrowing or swimming. Incision of the body-wall, which is necessary for cutting the ligaments, is followed by local contraction of the body-wall muscles that seals off the wound and prevents the normal functioning of that part of the body. Chapman (1951) found it extremely difficult to measure the coelomic hydrostatic pressure in Nephtys because of the great activity of the worms, and it has proved totally impracticable to do so when they are burrowing through the sand or swimming. However, a number of tentative conclusions may be drawn from a consideration of the structure of the ligaments, their disposition, and the behaviour of the worms.

The ligaments are composed of bundles of alternating inextensible fibrils and elastic filaments linked to cross-membranes which are also elastic and which partition the ligaments. The mechanical consequences of such a structure are these.

1. Since the elastic elements constitute less than half the total length of the ligament, and the rest is inextensible, the extensibility of the ligament is less than it would be were it composed entirely of the elastic material.

2. As a corollary of this, the restoring force generated by stretching the ligament a given amount is greater than it would be were the ligament composed entirely of the elastic material. (This may be readily verified by considering the expression for Young’s modulus, E=fLadL, where f is the force, a the cross-sectional area of the fibril, and L its length.)

Thus the ligaments appear to be suited for resisting deformation while permitting a degree of flexibility to the system, and for providing a restoring force when they are elongated slightly, without necessarily being under tension when no external forces are applied to the system. These conditions would not be met by wholly elastic or wholly inextensible ligaments, and the existence of ligaments with this unique construction in Nephtys suggests that the body of the worm is exposed to forces not experienced by other animals.

To judge by the disposition of the ligaments in the body of the worm, they have the following functions:

1. The median longitudinal ligament protects the nerve-cord from undue stretching when the worm is elongated. This is a function sometimes performed by a supra-neural muscle, as in nereids (Defretin, 1949; Smith, 1957). There is no such muscle in Nephtys.

2. The ligaments running from the attachment nodes on the ventral nervecord to those in the intersegmental groove prevent the deformation of the intersegmental body-wall during changes of shape of the segment. The chief antagonists of the longitudinal muscles are the dorso-ventral muscles, and the only muscles inserted in the intersegmental groove are those of the septum, the slender diagonal muscle (12), and the proximal posterior parapodial muscle (10). Some of the other extrinsic parapodial muscles may assist in this function, but they are inserted more distally in the parapodium and cannot serve to anchor this part of the body-wall. Several polychaetes have reduced circular muscles, though it is rare for them to be lacking altogether as they are in Nephtys, and it is significant that the intersegmental body-wall of nereids, in which the circular muscles are slight, is strengthened by circumferential fibres (Smith, 1957). We do not know whether Nephtys possesses similar fibres, but it is evident that in nereidiform polychaetes with a reduced circular muscle-layer, there is a need for additional support in the intersegmental region of the body-wall.

3. In worms in which changes of shape depend upon a hydrostatic skeleton, it is important that pressure changes produced by one set of body-wall muscles should be applied to the antagonistic set of muscles and not dissipated by the dilatation of other parts of the body-wall. None of the extrinsic parapodial muscles of Nephtys runs to the tip of the parapodium, an unusual situation in nereidiform polychaetes, and the ligaments inserted in the acicular and chaetel sacs may supplement the acicular muscles and the intrinsic parapodial muscles in preventing the protrusion of the parapodium when the body-wall muscles contract. The lateral longitudinal ligaments are disposed in such a way as to prevent dilatation of the base of the parapodium.

A possible consequence of the method of burrowing is that the impact of the proboscis against the substratum produces a high, transient, fluid pressure in the coelom of the worm. The peculiar elastic properties of the fibres, and their disposition to support the weakest parts of the body-wall and to brace the septa from the anterior side, are particularly suited to resist sudden shocks of this kind. While many other polychaetes burrow in sand, and some may use similar methods of burrowing to those employed by Nephtys, they nearly all inhabit permanent burrows. Nephtys is one of the few polychaetes that is able to live in shifting sands; it does not form a consolidated, permanent burrow, and it is predaceous. There is every likelihood therefore that burrowing is a much more frequent activity of this worm than of other worms living in sandy beaches and this may account for its unique structures.

We are greatly indebted to Professor G. Chapman for permitting us to quote some of his unpublished data on the mechanism of proboscis eversion in Nephtys. His (earlier) account of the method by which the proboscis is everted agrees with our own.

Chapman
,
G.
,
1950
. ‘
Of the movement of worms.’
J. exp. Biol
.,
27
,
29
.
Chapman
,
G.
,
1951
. ‘
Muscles of the proboscis of Nephtys hombergi and their use during burrowing.’
(Unpublished data
.)
Chapman
,
G.
,
1958
. ‘
The hydrostatic skeleton in the invertebrates.’
Biol. Rev
.,
33
,
338
.
Chapman
,
G.
and
Newell
,
G. E.
,
1947
. ‘
The role of the body fluid in relation to the movement of soft-bodied invertebrates. I. The burrowing of Arenicola.’
Proc. Roy. Soc. B
,
134
,
431
.
Clark
,
M. E.
, and
Clark
,
R. B.
,
1960
. ‘
The fine structure and histochemistry of the ligaments of Nephtys.’
Quart. J. micr. Sci
.,
101
,
133
.
Clark
,
R. B.
,
1956a
. ‘
The blood vascular system of Nephtys (Annelida, Polychaeta).’
Quart. J. micr. Sci
.,
97
,
235
.
Clark
,
R. B.
1956b
. ‘
The eyes and the photonegative behaviour of Nephtys (Annelida, Polychaeta).’
J. exp. Biol
.,
33
,
461
.
Clark
,
R. B.
1958
. ‘
The gross morphology of the anterior nervous system of Nephtys.’
Quart. J. micr. Sci
.,
99
,
205
.
Defretin
,
R.
,
1949
. ‘
Recherches sur la musculature des Néréidiens au cours de l’épitoquie, sur les glandes parapodiales, et sur la spermiogenèse.’
Ann. Inst. Océanogr. Monaco
,
24
,
117
.
Durchon
,
M.
,
1952
. ‘
Recherches expérimentales sur deux aspects de la reproduction chez les Annélides Polychètes: l’épitoquie et la stolonisation.’
Ann. Sci. Nat. Zool
.,
14
,
117
.
Ehlers
,
E.
,
1864-8
.
Die Borstenwiirmer, Leipzig (Engelmann)
.
Emery
,
C.
,
1887
. ‘
Intorno alla muscolatura liscia e striata della Nephthys scolopendroidesD. Ch.’
Mitt. zool. Stat. Neapel
,
7
,
371
.
Gray
,
J.
,
1939
. ‘
Studies in animal locomotion. VIII. The kinetics of locomotion of Nereis diversicolor.’
J. exp. Biol
.,
16
,
9
.
Gray
,
J.
and
Lissmann
,
H. W.
,
1938
. ‘
Studies in animal locomotion. VII. Locomotory reflexes in the earthworm.’
J. exp. Biol
.,
15
,
506
.
Langerhans
,
P.
,
1880
.
‘Die Wurmfauna von Madeira. IL’ Z. wiss. Zool
.,
33
,
267
.
Lillie
,
F. R.
, and
Just
,
E. E.
,
1913
. ‘
Breeding habits of the heteronereis form of Nereis limbata at Woods Hole, Mass.’
Biol. Bull. Woods Hole
,
24
,
147
.
Newell
,
G. E.
,
1950
. ‘
The role of coelomic fluid in the movements of earthworms.’
J. exp. Biol
.,
27
,
o
.
Schack
,
F.
,
1886
.
Anatomisch-histologische Untersuchung von Nephthys coeca Fabricius.Kiel (Zool. Inst
.).
Smith
,
J. E.
,
1957
. ‘
The nervous anatomy of the body segments of.nereid polychaetes.’
Phil. Trans. B
,
240
,
135
.
Snodgrass
,
R. E.
,
1938
. ‘
Evolution of the Annelida, Onychophora and Arthropoda.’
Smith- son. Mise. Coll
.,
97
(
6
),
I
.
Taylor
,
G.
,
1952
. ‘
Analysis of the swimming of long and narrow animals.’
Proc. Roy. Soc. A
,
214
,
158
.
Wells
,
G. P.
,
1944
. ‘
Mechanism of burrowing of Arenicola marina L.’ Nature
,
Lond
.,
155
,
396
.
Wells
,
G. P.
1948
. ‘
Thixotropy, and the mechanics of burrowing in the lugworm (Arenicola marinaL.).’
Ibid
.,
162
,
652
.
Wells
,
G. P.
1954
. ‘
The mechanism of proboscis movement in Arenicola.’
Quart. J. micr. Sci
.,
95
,
251
.
[Quarterly Journal of Microscopical Science
, Vol.
101
, part 2, pp.
149
76
,