Gulf and northern part of the North Island. The type locality of roseum is “Thames River in considerable depth.” Suter referred the common Hauraki Gulf shells to nucleogranosum, the type of which is South Australian; he differentiated this material from roseum apparently on the grounds of Quoy and Gaimard's original description of the latter species, which is translated directly into the Manual. That Suter was not personally acquainted with a species with “protoconch smooth; bristles of operculum simple” is suggested by the single specimen labelled roseum in his collection—“3726 Stephopoma roseum, Q. and G. 4 f. Rangitoto Channel.” This specimen is the basis of the sole record of roseum, apart from the type locality, in the Manual. The shell is worn and imperfect with the apex eroded. The whorls are smooth shouldered with a “Cyclostoma-like appearance,” but the specimen cannot be satisfactorily separated from the common Hauraki Gulf species, in which the shape of the whorls, particularly in beach-worn shells, is highly variable.

It is at least unlikely that the species described by Quoy and Gaimard as Vermetus roseus should not have turned up during the whole of the following century, and on reading the original account of this species, we may safely draw the conclusion that roseum and the neozelanic shells referred by Suter to nucleogranosum are one and the same species. Particularly is this view supported by the reference to “Thames River” as the type locality of roseum. The shells referred to nucleogranosum are common in the Hauraki Gulf, from which opens the Firth of Thames, and no other species of Stephopoma is upon record from this locality. In addition, the details of the animal given by Quoy and Gaimard are quite recognisably applicable to the Auckland species. Nor are the shell features taken by Suter from the original description sufficient to mark off two New Zealand species of Stephopoma, having regard to the different convention of description and illustration of a century ago. The rounded, non-carinate whorls might well belong to a specimen of the extremely variable Auckland species, in which, especially in the later whorls, the shoulder is often obsolete or wanting. Suter's statement in his key to the genus “protoconch smooth” is his only interpolation into Quoy and Gaimard's description, and is certainly not warranted from the material available to him. Quoy and Gaimard do not specifically mention the nucleus; though, as was long ago suggested by Morch, their reference to minute holes, as of some parasitic animal, may be due to a mistaken microscopic interpretation of the pustulations of the embryo shell, referred to below. A similar imperfection in microscopical technique was probably responsible for Quoy and Gaimard's figure of the opercular bristle, which is poor and oversimplified. The bristle represented is unlike that of any other Stephopoma, and cannot be safely relied upon. The colour description of roseum applies well to the Auckland species; unencrusted living shells are frequently flushed with pink or pinkish brown.

Thus the New Zealand nucleogranosum becomes Stephopoma roseum as the sole neozelanic representative of the genus. Verco's South Australian species, however, remains valid, by virtue of slight differences from the New Zealand shells, which were predicted by Finlay (1927). The writer is greatly indebted to Mr. B. C. Cotton,

of the South Australian Museum, for shell and operculum material of Stephopoma nucleogranosum. Roseum differs from nucleogranosum in the sculpture of the nucleus, which is ornamented with closely spaced, regularly arranged tubercles or pustules, relatively larger than the “numerous minute granules” of Verco's species. The pustules are arranged in rows along the growth lines, giving the nucleus a tessellated appearance back to the apical half whorl, which is a tiny smooth bulb. A row of 30–40 well-developed pustules runs round the periphery—a feature which Verco states to be quite wanting in the nucleogranosum. The aperture of the embryo shell is, as in nucleogranosum, trumpet-shaped, its rim projecting narrowly at the inception of the adult shell, which is thick and smooth, with fine accremental striae. Verco's figure of the opercular bristle is not quite like that of the South Australian specimen examined by the writer. In Milford roseum the opercular bristle differs only from that of nucleogranosum in the origin of asymmetrical lateral branches, which in roseum (Text Fig. 1) spring straight from the main shaft, but in nucleogranosum arise by the division of a single branch axis. This character may, however, be somewhat variable, and is probably not of constant taxonomic value.

Text Fig. 1—Stephopoma roseum Milford. A single opercular bristle from the outermost
annulus. Half the length of the long mediun seta has been left out. X

Dentition: Unknown in nucleogranosum; figured for roseum (Fig. 23). The central tooth has a short median cusp, flanked by four minute denticles on either side. The laterals are long and slender, strongly hooked antero-medially, the marginals curved and falciform, without serrations.

Localities for Stephopoma roseum: Thames River (type) Q. and G.; Rangitoto Channel in about 5 fathoms, Takapuna Reef, Bay of Islands (H.S.); Awanui Bay, North Auckland, embryos dredged in 12 fathoms (Finlay Collection, per W. la Roche); Milford Reef, under boulders at low spring tide; Otata Island, Noises Group, on vertical rock faces at low spring tide (J.E.M.).

The close resemblance of nucleogranosum to roseum is worthy of remark; when we consider also the fact that the New Zealand Pyxipoma is conspecific with the Australian P. weldii, it is evident that the present group of vermetids—unlike Serpulorbis—does not well exemplify the usually clear-cut disparity between Australian and