The fold which in S. nudus surrounds the mouth may be in the same species bent in so as to take the form of a double horse-shoe, the opening of which is always dorsal, just above the brain; in this case the mouth is crescentiform. In other genera the fold is broken up into discrete tentacles, and these are variously arranged; in Dendrostoma they are grouped together in four or six bundles round the mouth, but the more usual arrangement is the horse-shoe-like row of tentacles which overhang the crescentiform mouth, as in Phymosoma and some species of Aspidosiphon.
The ventral side of each tentacle is grooved and ciliated, and the grooves are continued into the ciliated mouth. Their dorsal surface is pigmented, and in the hollow of the horse-shoe lies a deeply pigmented epithelium covering the brain.
A blood-vessel courses up each tentacle, and usually two channels return the blood to the vascular ring which surrounds the mouth. In those forms which possess tentacles on the dorsal side of the mouth only, the ventral part of the vascular ring lies in the lower lip, which is tumid and swollen. The brain supplies a nerve to each tentacle.
When the introvert is retracted the tentacular ring is withdrawn and to some extent collapsed; in this condition it would be almost touching the rough external surface of the introvert. In some species of Phymosoma the delicate appendages of the head are guarded from the hooks on the introvert by a thin membrane or collar,[[479]] which completely ensheaths the retracted head.
Fig. 214.—A, Phymosoma granulatum F. S. Leuck. × 2. B, Head of the same. × 4. a, Pigmented pit leading to brain. The crescentiform mouth on the lower side of the figure is overhung by the tentacles.
When the introvert is fully extended the dorsal blood-vessel contracts and sends its blood forward into the vascular ring, and thence into the tentacles or tentacular fold, which are thus erected. In several species of Sipunculus, as S. nudus, S. norvegicus, S. robustus, S. tesselatus, there is a ventral blind tube as well as a dorsal, into which the blood is withdrawn when the head is retracted. In many other species in various genera, such as Phymosoma weldonii and Ph. asser, Dendrostoma signifer, S. vastus, the lumen of the dorsal vessel is increased by numerous hollow blind processes which it bears, hanging freely into the body-cavity. Three very small genera of Sipunculids—Onchnesoma, Petalostoma, and Tylosoma—are devoid of all trace of vascular system and of tentacles; the mouth opens in the centre of the anterior end of the introvert. In Onchnesoma the dorsal part of the lip is somewhat produced, so that the head has somewhat the shape of a Doge's cap, and in Petalostoma there are two leaf-like processes of the body-wall which guard the mouth.
The extent to which the intestine is coiled varies very much even in the same species; the axis of the coil is often supported by a spindle-muscle, but this is sometimes absent. The caecum, which opens into the rectum of S. nudus, is again a very variable structure, and when it is present varies remarkably in size.
The food of Sipunculids seems to consist almost entirely of sand, and their only nourishment must be such small microscopic organisms or particles of animal and vegetable débris as are to be found mixed with the sand. The alimentary canal is, as a rule, quite full of sand, and yet in spite of the tenuity of its walls they never seem to be ruptured. If the contents of the digestive tube be washed out with a pipette, it will be found that it requires considerable force to dislodge many of the sand-particles lying next the wall. These are more or less embedded in crypts or pockets of the wall, and as the sand passes along the intestine they probably serve as more or less fixed hard points, against which the sharp edges of the sand particles are worn off. Amongst the sand are usually to be found pieces of shell, sometimes with a diameter equal to that of the alimentary canal; these are usually rounded, but their angles may have been removed by attrition before they entered the mouth of the Sipunculid.
In S. tesselatus the sand is to some extent held together by a mucous deposit; in those cases where there is no sand in the intestine, there is always a coagulum of mucus, and the walls are contracted and thick; when full of sand the walls are tensely stretched and very thin. This thinness of the wall of the alimentary canal seems ill-adapted to a diet of sand, nevertheless it is also met with in other great sand-eating groups of animals, such as the Echinids and the Holothurians.