General Appearance.—Shell either extremely much depressed and irregular, or globulo-conical, or more rarely cylindrical and elongated. Walls, either with a few rather broad, smooth, irregular, longitudinal folds, or, in var. sordidus, with numerous sharper and more prominent longitudinal ridges; basal margin very sinuous. Colour, either fine rich peach-blossom purple, or so pale as to be almost white; or in var. sordidus dirty white, generally stained greenish from confervoid matter. Orifice small, oval, entire. Radii very narrow, white, or not at all developed, and with even the sutures not distinguishable. The purple coloured varieties, with the narrow white radii, the small oval orifice, and folded walls, have a very pretty appearance, which is far from the case with var. sordidus. The largest specimens attained a basal diameter of .6 of an inch, but this is an unusual size: I have seen a cylindrical specimen of var. sordidus one inch in length.

The opercular valves are united to the sheath by unusually strong membrane: internally, their upper parts are stained purple. Scuta, these vary considerably in breadth, some being even broader than in Pl. 8, fig. 5 c, and others as much elongated as in fig. 5 d: these latter come from an elongated cylindrical shell. The valve externally is unusually convex: the apex, also, projects freely to an unusual degree. Internally, the articular ridge is moderately prominent: the adductor ridge is prominent and much curved: in the upper part it either lies close to, or at some little distance from the articular ridge. The lateral depressores are attached to several little crests, occupying a cavity, (often bordered above by a little ridge) close beneath the adductor ridge. Tergum, extraordinarily narrow and elongated; beak triangular, purple: a longitudinal furrow runs down the valve: spur narrow, long, bluntly pointed, lying close to the basi-scutal angle of the valve: the scutal margin is nearly straight and parallel to the spur. Internally, the articular ridge is prominent: the crests for the depressores are moderately well developed: the upper part of the valve is marked by a purple patch, bounded on one side by the articular ridge, and on the other side by a very slight special ridge. There is some variability in the narrowness of the whole valve, and in the sharpness of the spur.

Parietes.—The under surface, in the more depressed varieties, is roughened in a remarkable manner nearly or quite up to the sheath, with very irregular, projecting, and branching ridges, and sometimes with depending points. These ridges and points are granulated on their under surfaces. The roughened surface in the more conical varieties is confined to the basal inner margin. This structure is nearly the same as that in B. imperator, represented (Pl. 8, fig. 4 c), but the little ridges are here more apt to be concentric, instead of radiating. The lateral edges of the compartments on the inside, especially the carinal edges, project inwards beyond the inner surface of the shell. The sheath is but little hollow on its under side. The diametric growth of the shell seems to be quite capricious; in the same group, some individuals thus increasing, and others not at all. When the radii are developed they are narrow and white, with their upper margins nearly if not quite parallel to the basis: their sutural edges are formed by large, irregular, branching teeth or septa. The alæ, also, have their sutural edges coarsely crenated; and when the shell increases by diametric growth, they are added to above the opercular membrane, so as to be nearly square at top.

Basis.—When a shell is removed from the surface of attachment, and inspected even under a lens, there appears to be no basis whatever, either adhering to the shell, or to the supporting surface: but when a more careful examination is made, with a higher power, an excessively thin, translucent, calcareous, irregular layer, or rather film, can almost always be discovered. This would be more easily distinguished if the specimens had adhered to rock instead of to the rugged shells of molluscs. At one time I thought the basis was partially membranous, for I have certainly scraped off small fragments of membrane from the supporting surface; but these, when examined under the compound microscope, seemed always to consist of a thin sheet of the yellow cementing tissue; and in some instances, a brittle film of shell, representing the true basis of the cirripede, still adhered to the upper surface of the membranous bits of cement. Nevertheless, so imperfect is the calcareous basis, that I should not be surprised if portions of a true membranous basis did really in some cases exist.

Mouth: labrum with the notch wide, generally with a few little teeth; mandibles with three teeth, and some inferior knobs; maxillæ notched. Cirri, first pair with one ramus shorter by three segments than the other ramus. Second and third pairs short, of nearly equal length, thickly clothed with spines; segments very little protuberant. Posterior cirri, having elongated segments, supporting seven pairs of spines.

Var. sordidus.—This form is very common on the tidal shores of the Strait of Magellan, and of the southernmost parts of Tierra del Fuego, near Cape Horn: it lives attached to rocks, mytili, and logs of wood, and is associated with Chthamalus scabrosus. It almost certainly is the most antarctic form of the genus Balanus. If I were guided by external appearance alone, I should certainly separate this form specifically from B. flosculus, but, as will be seen in the following description, the differences consist only in var. sordidus being much duller and rather differently coloured, in the longitudinal folds being sharper and more prominent, and in the whole shell being rather more globular, and on an average rather larger; but in the true B. flosculus there is considerable variation in all these respects, as there likewise is in var. sordidus; thus some of the cylindrical varieties of the latter have less prominent ridges than even B. flosculus. In general appearance I have seen some nearly, but not exactly, intermediate forms; therefore, I do not feel positive that these forms may not be specifically distinct, but have failed, after careful examination, to find any sufficient diagnostic characters. Moreover, in the case of Balanus lævis, I was led to believe that there is an equal and somewhat analogous amount of variation in the specimens inhabiting Tierra del Fuego and northern Chile; and in this case I was enabled to show the existence of strictly intermediate forms in the intermediate districts.

The shell in var. sordidus is generally globulo-conical, dirty white, frequently with a green tinge, from the growth of confervoid matter. Orifice small. The exterior surface is covered with numerous prominent, narrow, sharp ribs or folds, the basal margin being serrated with projecting points where the folds terminate. When the radii are not developed, the sutures are very often obscure. Internally, the shell is faintly tinted of a port-wine purple. In all points of structure this form is identical with the true B. flosculus. In some few specimens the whole exterior surface was disintegrated and smooth; and this is generally the case with the upper parts of the shell. Some other specimens, which had grown crowded together on wood, had become cylindrical, and consequently the orifice was as large in diameter as the shell, namely, half an inch: in some of these cylindrical varieties the sheath was entirely soldered to the walls. The largest specimens which I have seen were .6 of an inch in diameter; and above one inch in height.

Affinities.—This species, in its opercular valves, even in such trifling characters as the strength of the opercular membrane, and in its cirri, approaches closely to B. cariosus. We even see on the under side of the scutum, in that species, a single little ridge, foreshadowing, as it were, the crests for the lateral scutal depressores, so remarkable in our present species. In the structure of the shell and of the basis, B. flosculus is much more closely related to the last species, or B. imperator. If it had been possible to have arranged the species in a single line, B. flosculus ought undoubtedly to have been placed between B. cariosus and imperator.