There is not, by any means, the same amount of diversity in the structure of the shell in the class of Gasteropoda, as that which exists among the several tribes of Conchifera ; a certain typical plan of con struction being coinmon to by far the greater number of them. The small proportion of animal matter contained in most of these shells is a very marked feature in their cha racter ; and it serves to render other features indistinct, since the residuum left after the removal of the calcareous matter is usually so imperfect, as to give no clue whatever to the explanation of the appearances shown by sec tions. Nevertheless, the structure of these shells is by no means homogeneous, but al ways exhibits indications, more or less clear, of an original organic arrangement. The poreellanous shells, as formerly stated (vol. ii. p. 384), are composed of three layers, all pre senting the same kind of structure, but each differing from the others in the mode in which this is arranged. This structure was de scribed by Air. Gray 1- as the result of rhom boidal crystallisation ; each layer being com posed of thin laminm placed side hy side, which separate from one another in the planes of cleavage when the shell is fractured. As first pointed out, however, by Mr. Bowerbank, each of these laminee really consists of a series of cells in close apposition ; and the plates are disposed alternately in contrary directions, so that each series of cells intersects the one beneath it nearly at right angles, as seen in fig. 420. Although the intimate structure of each of the three layers of the shell is essen tially the same, yet the disposition of the larninm is not the same in any two adjoining ones,—an arrangement which adds greatly to the strength of the shell. The planes of the laminm are always as nearly as possible either parallel or at right angles to the lines of growth ; those of the inner and outer layers always having the same direction with each other, but those of the middle layer being set at right angles to them. When, therefore, a section is made parallel to the surface of the shell, it will cut the edges of the laminm of which the layers traversed by it are composed; but if the section be made in a direction per pendicular to the surface, and pass through the middle layer in the plane of its lamince, it will cut through the edges of the laminm making up the interior and exterior layers ; whilst if the section traverse the two latter in the plane of their laminx, it will cut across the laminae of the middle layer.
The principal departures from this plan of structure are seen in Patella, Chiton, Ha hotis, and Turbo and its allies. In Patella, the inner and outer layers are composed of large and irregular laminm, by no means firmly adherent to one another ; but the middle layer is made up of tolerably regular polygonal cells, which form only a thin stratum in some parts, whilst in others they are elongated into prismatic cells; and the directions of the laminm, of which the inner and outer layers are composed, instead of being conformable with each other, are at right angles. In Chiton, the external layer, which seems to be of a delicate fibrous tex ture, but which is of extreme density, is per forated by large canals, which pass down obliquely into its substance, without pene trating, however, as far the middle layer. The middle layer, as in Patella is distinctly cellular • whilst the internal h'as the same nearly-h'omogeneous texture as the external, but shows no trace of perforations. The
peculiarities of structure presented by lia liotis have been already described. In Turbo and its allies, the inner layer is nacreous, and the middle one is made up of large cells : the cellular structure is also very evident in the solid operculum of Turbo, when reduced to sufficient thinness.
That the shell-substance in Gasteropoda is formed in the first instance by. the agency of cells, however indistinct thetr traces may subsequently become, is further apparent from the researches of Mr. Bowerbank on the growth of the shell of the common garden-snail (Helix aspersa); and his obser vations further confirm the opinion already expressed, that the formation of each layer of shell is a progressive operation ; new matter being added to its interior after the exterior has been consolidated.
Passing by the Pteropoda, whose delicate membranous shells present no very distinct structure, we come to the testaceous Cephalo poda, of which there are but few species now existing. The shell of Nautilus pompilius bears more resemblance to that of bivalves in its intimate structure, than to that of the Gasteropodous univalves ; the three layers of perpendicular laminm, so characteristic of the latter, not making their appearance here ; and of the two layers of which the shell is com posed, the inner one being nacreous, whilst the outer one is made up of an aggregation of cells of various sizes, those which are nearest the external surface being generally the largest. In the thin shell of Argonauta, the same kind of irregular cellular structure can be easily distinguished, as in the outer layer of the shell of IV autilus ; but there would seem to be nothing comparable to the inner layer of the latter. The shell of Spi rula must be cons.dered to bear a greater re semblance, as regards its relation to the ani mal, to the Sepiostaire of the Cuttle-fish, than to the chambered shell of the Nautilus ; although it so closely approximates the latter in its own conformation. This being the case, it is interesting to find that the intimate structure of the shell has a much greater resemblance to the Sepiostaire than would be supposed from its general aspect. For al though its texture seems uniform, and its minute parts are composed of an aggregation of calcified cells, yet its surface is marked by sinuous lines, closely resembling those which are seen upon the transverse plates of the Sepiostaire ; and these lines or bands project in such a degree, that they might be con sidered as rudiments of the vertical partitions which connect these plates. The Sepiostaire having been formerly described in some detail (vol. i., p. 546), it will only be re quisite here to mention, that the calcified layers which alternate mith horny membranes to form the shallow coDe or cup, exhibit a distinct cellular structure, when the section is made sufficiently thin; and that indications of a similar structure may also be perceived in the delicate and fragile plates which are arranged obliquely upon one another in the hollow of *this cup. Few of the numerous fossil shells referable to this class have yet been exatnined ; it may, however, be stated as an interestino. result of microscopic ob servation, that the " spathose guard" of the Belemnite is thereby proved to be composed of long prismatic cells, radiating from the centre to the circumference; closely re sembling in their general arrangement those of the massive tube of Septaria gigantea, the great sand-boring Teredo of Sumatra.