indeed, between the operculum and the valve of a shell consists in the former having no cavity, the cone of which it is formed being either very much depressed, so as to become nearly flat, or even concave, as in the annular or some subannular opercula, or very much compressed, forming only a spiral riband, as in the spiral ones. Opercula are never attached to their shells by ligaments or by any other means than that of the adductor muscle; and they are always free, except in the genus Capulus, which offers a remarkable anomaly in this part being immediately attached by its outer surface to other marine bodies.

It is proper to observe, that in describing the operculum I have called that the anterior margin, which is nearest to the pillar of the shell, as when the animal is walking this part is directed towards its head; and the right and left extremities are the parts corresponding with the right and left sides of the body. When the operculum is placed in this position the left end is that which fits into the front, and the right that which corresponds to the hinder part, of the mouth. This distinction of parts may, at first sight, be thought trifling; but to the practical zoologist its importance is very great. The position of the nucleus of the operculum is, for example, almost the only conchological character by which four distinct genera of shells can be divided into groups, although their animals differ both in external form and in internal anatomy. In the genera Bithynia and Paludina, which have the nucleus of the operculum nearly central, the animal has short tentacles and no air bag, whilst in Ampullaria and Ceratodes, which have the nucleus of the operculum on its anterior side, the animals have very long tentacles, and a large air pouch by the side of the branchiæ.

Fig. 78.

Opercula may be divided into three very distinct kinds, according to their form and manner of growth, and these may be again subdivided according to the mode in which they are covered with various deposits. The annular operculum (Fig. 78) may be considered the most simple, the very depressed cone of which it is formed being nearly regular, with the apex more or less central, and the coat of new matter, by which it is increased in size, forming complete rings round its circumference; in which particulars it may be compared to the simple conical shells of the genera Patella, Fissurella, &c. This kind of operculum does not alter its place in the mouth,

and the muscle of attachment only moves nearer towards its anterior edge as the addition of new matter on that side renders such a displacement necessary in order to keep the muscle, which enlarges in proportion, in its proper situation with regard to the pillar of the shell.

The subannular opercula (Fig. 79) may be regarded as intermediate between the annular and spiral forms, partially combining the characters of each; but I think it better to consider them apart from the others, inasmuch as they are peculiar to those animals of the Ctenobranchous Mollusca, which are provided with a siphon in front of the mantle for conducting the water to their branchiæ, such as the Murices, Buccina, Strombi, Melania, Melanopsides, Aulodi, and the anomalous genus Phorus. They are all of

a horny texture, and are characterised by their very depressed cone being somewhat oblique, with its nucleus placed at or near the left end, and the lines of growth forming more or less complete rings around it, but always becoming wider apart from each other as they approach the right side. The left end, towards which the nucleus is placed, is generally acute, and the opposite extremity rounded, which is just the reverse of what takes place in the annular opercula, where the right end is acute and furnished with a fold proceeding from the nucleus, and the left side is rounded and broad. In most of the opercula of this division the muscular scar occupies the greater part of the internal surface; is marked with more or less regular

concentric rings; and is surrounded by a thickened callous deposit, which is broadest on the outer side (Fig. 79, b). This scar appears gradually to approximate towards the right side of the operculum, the part left free, as the scar advances, becoming covered by the callous deposit. The lines on the scar appear to be very constant in the various species, but they are sometimes distorted, and form several centres instead of one. They have no relation to the rings of growth on the outer surface, but have generally a centre of their own, placed at some distance from the left end of the operculum; and they appear to be formed by the successive additions made to the edge of the adductor muscle, which is marked with lines resembling those on the The greater number of these opercula do not alter their relative position in the mouth of the shell; but a few, like those of Fusus fornicatus, in which the end of the cone is slightly curved, move during their lives perhaps to the extent of a quarter or half a turn on the end of the muscle. In some instances, as in the operculum of Strombus, where the foot of the animal is very small, the apex or left extremity is obliquely elevated and free: as the rings of growth are added, this extremity elongates, and acquires such a

scar.

resemblance in form to the claw of an animal, as to have induced the ancients to call such opercula Elks' hoofs (Fig. 80). The scar of attachment is small in comparison with

the size of the operculum, and is situated on its right side: it is cordate and marked with oblique rugose grooves, and with a strong central ridge, which is continued in the form of a rib down the middle of the under side of the free part of the operculum. In some species the posterior edge of the operculum is serrated.

In both the annular and subannular divisions, the disk to which the operculum adheres is formed entirely of the muscle of attachment, and of a membranaceous fringe by which its edge is surrounded. This fringe is free from the back of the foot and is widest posteriorly: it doubtless secretes the coat of the layer of growth, and the deposit which borders the edge of the scar, whilst the muscle itself is provided with the means of secreting the proper coat of the scar.

The spiral opercula are so called because the elongated, cartilaginous or shelly, compressed, riband-shaped cone of which they are formed is twisted into a spire, of few or many whorls (Fig. 81). In these, the new matter by which

they increase in size is deposited only on the extremity of the last whorl, which in spiral opercula may be regarded as analogous to the mouth of spiral shells, as the entire circumference of the annular opercula, with their subcentral apex, is analogous to the edge of the base of a Patella. The lines of growth are marked by curved concentric lines extending transversely across the whorls. Opercula of this character vary much in the number of their whorls: when the whorls are few in number the cone increases rapidly in size, and the operculum is of an ovate shape, with the nucleus approximated to the left extremity; when the whorls are more numerous, the cone increases in size more gradually, the nucleus is central, and the operculum is more orbicular. In all these cases the edge of the extremity of the last whorl is constantly anterior, that is to say, when the operculum is in the mouth of the shell, this part of it is directed towards the inner lip; when the operculum is ovate and of few whorls

it generally occupies the whole length of that lip; but in the orbicular and many-whorled opercula, it is directed towards the hinder part of the inner lip, near the angle formed by its junction with the outer. Now every time that the animal adds a new layer on the end of the last whorl, the operculum, to allow of this part continuing to occupy the same position, must make a slight turn backwards on its centre, which is the nucleus of the spire, whether the spire be placed either towards one end or in the centre of the operculum. This rotation on the adductor muscle, although it may at first sight appear improbable, bears a striking analogy to several other phenomena of the same kind which are continually taking place in the animal economy: I need only add two well-known and apposite examples :1st, the gradual change of situation of the adductor muscle as it passes down the pillar of spiral shells, which in some of the long turreted species, such as Turritella archimedes, where there are as many as thirty whorls, must have been carried to the extent of thirty complete revolutions on this part; and, 2nd, the change of place of the adductor muscle in bivalve shells. In the opercula of the Littorinæ and Natica, which consist of a few very rapidly enlarging whorls, the motion and consequent alteration of position of the place of attachment is very gradual and slow; but in those orbicular opercula which are composed of many gradually enlarging whorls, as in the Trochi and Monodontæ, the place of attachment must be continually changing, as many complete revolutions being made as there are whorls in the operculum. These are sometimes extremely numerous: in a small specimen of Turbo pica now before me, there are seventeen or eighteen, and in some Trochi I have counted still more. In the spiral opercula, which thus rotate on their axis, the nucleus, which is the centre of motion, is always included in the scar, and adherent to the muscle of attachment: it is often furnished internally with a small spiral process, buried in the muscle, and resembling the end of a screw. On the contrary, in the annular and subannular opercula, which have no rotatory motion, the nucleus is often removed from the point of attachment, as is well illustrated in those of the Strombi.

In many of those ovate or suborbicular spiral opercula which are formed of a few rapidly enlarging whorls, as, for example, those of the genus Littorina, the inner surface exhibits the lines of growth as well as the outer, there being in these cases no internal deposit. In these opercula the adductor muscle is anterior, and occupies more than half the