In those Lamellibranchiata which possess siphons, the pallial line is incurved beneath the impression of the posterior adductor muscles, forming a bay or sinus. Such shells are said to be sinupalliate.

The epicuticula varies in thickness in different Lamellibranchiata, and it is sometimes complicated in structure as in Mytilus edulis. It is rather thin in Anodonta. The prismatic layer consists of more than one layer of calcareous prisms, which vary in size and external shape. The different layers of prisms are held together by remains of the organic substance (conchiolin) in which the calcareous prisms are deposited. At their first appearance these bodies are, in Anodonta at least, more or less rounded and irregular in shape, and at some distance apart. They increase in size, and at the same time new prisms appear between those first formed. In some instances, e. g. Cyprina islandica, the prismatic layer consists chiefly of a mass of granular calcareous matter: in other instances it is very distinctly lamellate, like the nacre, e. g. Astarte borealis. It may closely resemble the Gastropod shell, and consist of vertical laminae, e. g. Cardium. The nacre consists of alternating lamellae of conchiolin, and of conchiolin containing calcareous deposits, which appear exactly in the same manner as the prisms of the prismatic layer.

In the region of the various muscular impressions it is 'transparent,' and here the muscular fibres are directly continuous with the shell-substance, and not separated from it by an epidermis. According to F. Miiller, the inner surface of the nacre is covered by a soft layer.

The Cyclas cornea of the freshwaters differs from other Lamellibranchiata in the structure of its shell. There is no prismatic layer, and the organic portion of the nacreous layer is reticulate. Short wide unbranched canals extend into the substance of the shell and are lined apparently by extensions of the mantle. In this point they differ from the branching canals so often found in the substance of the shell in other Lamellibranchiata, which, according to Kolliker, are due to algal parasites. Ehrenbaum, however, who mentions them in various genera, says nothing as to their contents.

The structure of the calcareous parts of the Lamellibranch shell does not appear to be by any means so uniform as is often supposed.

The ligament in Anodonta consists of an outer and inner part. The outer is laminated and passes gradually into the epicuticula. The inner is striated radially, and consists of radial fibre composed of two different alternating substances refracting light differently. Hence in cross sections this inner part appears to be laminated like the outer: but the mode in which it splits proves its real radial structure. It must probably be regarded as continuous with the nacreous layer. Where it borders the nacreous layer internally there is a distinct nacreous ridge, the homologue in Anodonta of the region which is produced into teeth in Unio, etc, and, like the teeth, giving attachment to muscles ascending from the foot (F. Muller). The border of the ligament itself is connected with numerous isolated bundles of muscles; and just as the 'transparent' portion of the nacre in the muscular impressions is continuous with muscle fibres, so here the fibres of the inner part of the ligament are continuous with muscle fibres.

The relations of the ligament to the shell-valves show that, strictly speaking, the valves ought to be regarded as parts of a continuous structure. The dorsal region of this structure does not undergo calcification, or only to a very slight extent, inasmuch as the economy of the animal requires that it should remain flexible. It is an adaptation of an originally univalve shell.

The whole shell is generally regarded as a cuticular secretion, the cells near the edge of the mantle forming the epicuticula; those of the part a little remote, the prismatic layer; and those of the general surface the nacre. Tullberg appears to take the view that the organic part of the shell is produced by a fibrillation of the cells. F. Muller believes that it increases by intus-susception, and states that in Anodonta the surface of the mantle is separated from the shell over a large space extending from the pallial line and adductor muscles as far as the attachment of the muscles to the ridge bordering the ligament (supra). Judging from analogy it is more reasonable to class the shell as a cuticular formation. This conclusion is borne out by Osborn's experiments on the Oyster. He studied the formation of shell by snipping away portions of the already formed shell, and placing on the exposed surface of the mantle a disc of thin glass such as is used in microscopic work. He found that a gummy film was formed on the surface of the glass by the columnar cells of the mantle, which hardened in forty-eight hours into a tough leathery membrane. Crystals of lime, for the most part scaly, appeared in this membrane, which became stony in six days.

It is probable that the gummy films formed by the mantle-cells are charged with lime carbonate, which eventually crystallises, its crystalline form being modified by the organic matter of the films, as is the case with crystalline products formed in the presence of colloid matter.

The calcareous substance of the shell is chiefly composed of Calcium carbonate. Traces of Calcium phosphate, silica, alumina are sometimes found. The carbonate of lime is sometimes in the form of Calc-spar, e. g. Ostrea, Pecten, and such shells retain their integrity when fossilised. It is sometimes in the form of Arragonite, in the nacreous layer only, e.g. Pinna, or in the prismatic also, as in a very large number (Sorby). The Arragonite usually dissolves away when the shell is fossilised, and then either the inner layer only or both layers are lost as the case may be, leaving a stony nucleus or cast.

The thickness of the shell does not depend upon the amount of lime in the waters in which the animal dwells, but rather on the workings of its tissues, modified by surrounding influences, whether chemical or non-chemical. This may be readily seen by a comparison of the dense shell of a Pearl Mussel (Unio mar-garitifer), from the mountain-streams of Westmoreland, with the thin shell of Anodonta from Oxford waters so much richer in lime.

Structure of the shell. Ehrenbaum, Z. W. Z. xli. 1884; F. Muller, Z. A. viii. 1885; in Anodonta, F. Muller in Schneider's Zoologische Beitrage, Breslau, i. pt. 3, 1885; cf. Bronn, Klass. und Ordnungen des Thierreichs, iii. 1, p. 330: and Sorby, Presidential Address, Geological Society's Journal, xxxv. 1879; of Cyclas cornea, Leydig, Archiv f. Anat. und Physiol. 1855, and F. Muller, op. cit. supra.

Hinge teeth. Neumayr, SB. Akad. Wien. lxxxviii. Abth. i. 1883.

Formation of shell in Oyster. Osborn, Biological Studies from the Laboratory of Johns Hopkins University, ii. 1882; cf. on Embryo Oyster, Ryder in Bull. U. S. Fish Commission, ii. 1883. p. 383.

On 'Molecular Coalescence' and on the effect of Colloids upon the form of Inorganic Matter, Ord, Q. J. M. xii. 1872, and St. Thomas's Hospital Reports (N. S.), ii. 1871.

Formation of Pearls.Pagenstecher, Z. W. Z. ix. 1858, von Hessling, Die Perlmuscheln und ihre Perlen, Leipzig 1859, and in Z. W. Z. ix. 1858, p. 453.

Composition of shell in relation to lime in water, etc. Voit, Z. W. Z. x. 1859-60. Vegetable parasites in shell, etc. Kolliker, Z. W. Z. x. 1859-60.