the posterior extremities are expanded, the siphons (Fig. 10, e and d) can be carried outside into the water; at the same time the access of the two siphons, and consequently the entrance and exit of water, can be more or less hindered. As we have seen, the mantle is prolonged in the direction of the shell in an appendage which extends over the two sides of the dorsal surface of the valves (Fig. 1, a, and Fig. 2, a), the central portion of which forms a swelling of considerable thickness, composed of various anatomical elements; beneath the epidermis the tissues are partly of vesicular and partly of membranous character, which, through inherent powers of swelling and hardening under the action of the blood, serve a purpose in operating the movement of the valves.
To explain the physiological rôle of this organ (a, Figs. 1 and 2), it is necessary to recall the fact that it receives on either side, in the arched folds of the mantle, the neck-portions of the valves of the shells. By the contraction of the bundles of muscular fibres, the two valves would separate slightly one from the other, a movement which is still better understood if it is proved that that organ can become hard by the afflux of the blood and thus furnish a better fulcrum for the action of the muscles. Up to a certain point this part is similar to the hinge-ligament of other bivalve mollusks; but only in this respect, that it serves to open the shell. For the true ligament, wherever it exists, is always composed of elastic tissue, and its action is purely passive, while with the teredo the opening of the shell is a muscular action and consequently active. Moreover, the hinge is wanting in this case, which allows the supposition that the animal has the power of modifying at will, by the partial contraction of its muscles, the direction in which the valves separate, so that it may be at one time the middle parts and at another the anterior parts of the valves which separate most from each other. Besides, the effort which this action demands is extremely feeble, and the movement of the valves themselves is very limited.
There are two adductor muscles. The first and largest is already well known; it has been described by all writers who have made the teredo an object of study. It extends (Fig. 11, m) between the two valves in the form of a muscular mass, relatively quite large, and occupies about two-thirds of the length of the shell and one-third of its width. It rests on either side on a sort of pad situated at the limit between the middle and neck parts of the valve. The second or small adductor muscle, which appears to have escaped the attention of most observers, is found near the dorsal side of the shell in the cavity between the anterior portions of the valves. One can see its exterior surface, clothed with a thin epidermis and slightly projecting immediately in front of the pallial prolongation, which extends over the dorsal face of the shell; in appearance it is only a continuation of this muscle, but in reality it is entirely distinct.
The principal mass of this muscle is implanted upon the sides, bent