other vermetids, possessing a sharp, upraised spiral lamella of two whorls. Petaloconchus, although related to the above genera, has several other features setting it some distance apart. In the radula, the postero-lateral processes of the central tooth tend to be long and exaggerated. The female has no median pallial slit for the attachment of the egg capsules, which are retained in a cluster filling the pallial cavity until the emergence of the embryos. In the visceral mass, the anterior lobe of the digestive gland and its diverticulum from the gut have disappeared; the stomach otherwise resembles that of Novastoa.

Somewhat separated from the above genera, and related closely among themselves, are the species of Serpulorbis and Aletes, with which Bivonia is probably associated. This is the mucus-feeding branch of the family. Evolution has evidently proceeded furthest in such forms as Serpulorbis gigas (Yonge and Iles, 1939) and in Aletes. Serpulorbis zelandicus shows a transitional stage with the power of ciliary feeding still well developed (Morton, 1951). Bivonia triqueter—of which the only detailed account appears to be Lacaze-Duthier's memoir (1860)—is stated by Yonge to be a mucus trap feeder. Aletes squamigerus according to MacGinitie and MacGinitie, 1949, “secretes mucus that extends upward in the water as a triangular sheet. This sheet of mucus is allowed to float and wave in the water for a while, then the animal pulls it down and eats it, with what food material has adhered to it. When (the animals) occur in clusters, the fan-shaped sheets of mucus they put out become entangled, and the table with its bill of fare becomes a community affair. When one member in such a group begins to eat the mucus sheet all the others start swallowing. The sheets of mucus may extend upward into the water for five or six inches, and the upper edges may be somewhat frayed.”

The radulae of Serpulorbis zelandicus (see Morton, 1951), Serpulorbis gigas and Aletes squamigerus (Text Fig. 1, d), the type species of Aletes, are only with difficulty distinguishable. From data supplied by Dr. Keen, the apex of Aletes is almost exactly like that described for Serpulorbis zelandicus (Morton, 1951). The nucleus is three-whorled, translucent white, and the pitted sculpture of the first two or three adult coils is alike in both genera. Both groups have lost the operculum in the adult, while Bivonia retains it as a tiny vestige at the centre of the foot. Since adult sculpture is unreliable, it may be doubted whether there are any features accessible to the conchologist whereby Aletes and Serpulorbis may be safely separated. In the animal of the various species there may be quite considerable adaptive differences—for example, in the relative development of the pallial organs of Serpulorbis zelandicus and Serpulorbis gigas.

In the Vermetidae especially, an appreciation of the structure and biology of the animal is necessary for a proper understanding of the group. For taxonomic convenience it is desirable, however, that the final arbiter in separating genera should be a conchological character. Radulae tend to be rather conservative as between genera; perhaps their most useful variation is in the shape of the central tooth. Opercula, when present, form a reliable natural character. The most valuable evidence in determining genera will probably be yielded by the nuclear form and sculpture; Dr. Keen is at present completing