Neogastropods are usualiy accepted as the most advanccd prosobranchs, though their organization is approached in several respects in some higher families of Mesogastropoda. This seems, however, to be due to parallel evolution and the neogastropods originated from a much lower grade of mesogastropod. Although some workers derive them from an archaeogastropod stock there are too many features in their anatomy characteristic of mesogastropods rather than of archaeogastropods for this to bc acceptable. On the whofe, neogastropods are a rather uniform group of prosobranchs in their shell, external features, and internal anatomy. In only one System do they show, by comprison with archaeo- and mesogastropods, both extreme specialization and considerable variation: this is the gut, which is in several ways unlike that of any other prosobranch. This is to be associated with their carnivorous way of life, in which respect they again differ markedly from meso- and archaeogastropods. Taylor, Morris Br Taylor (1980) have shown how neogastropod species differ amongst themselves not, primarily, in their rnode of life, but in their often narrow choice of prey. Since the anatomical requirements for predation are more or less constant, the different species remain similar in organization and are often sympatric. In these respects neogastropods differ markedly from mesogastropods, whose adaptive radiation has been extensive and primarily in relation to mode of life. Separation of neogastropods from mesogastropods rests mainly on the siphonal canal in the shell, the siphon on the mantle edge, the rachiglossate or toxoglossate radula, and the presence of a pleurembolic proboscis or one of its varieties (Smith, 1967). The osphradium is large and its axis carries a double series of lamellae, giving it a gill-like appearance. Males always have a penis and females usually a ventral pedal gland. lnternally the anterior part of the alimentary caiial has becorne elaborate, with a complex glandular equipment, and the wall of the kidney is more folded than in mesogastropods. The nervous systern is concentrated, though the visceral ganglia remain posteriorly placed. Eggs are laid in capsules attached to the substratum. A free larval stage is often suppressed and food eggs are common, but neither of these features has much taxonomic significance, occurring apparently randomly throughout the group. Because of their general similarity classification of the Neogastropoda has proved to be no easy task, and there is still no universally-accepted subdivision of the order into superfamilies. It is generally agreed, however, that the order may be split into two groups, primarily on the basis of radular structure. The more primitive of these, the Rachiglossa, has a radula with typically 3 teeth per row; the more advanced, the Toxoglossa, has a radula which, in more primitive genera, resembles the rachiglossate, but which Comes, in more advanced toxoglossans, to have only a single tooth in action at a time. Each tooth has then become scroll-like and is used for the injection of poison from a poison gland into the prey (Shimek & Kohn, 1981). The group Toxoglossa is agreed to contain the superfamily Conacea which includes (as Recent forms) the families Turridae, Conidae, and Terebridae, all with poison apparatus, though with very different shells. Risbec (1955), followed by Taylor & Sohl (1962), has added a second superfamily Mitracea containing, in the family Mitridae, a grouping of genera selected from that family as earlier understood. These have a rachiglossate radula and an apparent poison gland not irnrnediately comparable with that of undoubted toxoglossans. This reclassification of mitrids has not found favour with subsequent workers (Cernohorsky, 1966, 1970; Ponder, 1972). Ponder (1973) made a case for adding a third suborder to the two mentioncd above. This was to contain the single superfamily Cancellariacea with the one family Cancellariidae. The case rests on the unique character of their radula. It is, however, when one turns to the remaining rachiglossan families and- attempts to assign them to superfamilies that difficulties mount. Three groupings Iiave been conventionally recognized - Muricacea, Buccinacea, and Volutacea, though it has often appeared that the last was a collection of animals not obviously assignable to the other two rather than clearly related amongst thernselves. Ponder (1973) came to the somewhat pessimistic conclusion that all rachiglossans should be put into a single taxon, for which he used the name Muricacea. It seems to us, however, that certainly within the limited group of anirnals with which we have to deal here, but even in a broader context, there is still some validity - and certainly convenience - in the older Separation, when due importance is given to internal anatomy; we propose, therefore, to retain the three superfamilies in dealing with a group which is otherwise too large for easy treatment. We adopt this arrangement the more readily as we have no volutacean mernbers of the fauna with which we have to deal, provided that we accept Ponder's proposal to create a separate superfamily for cancellariids. This allows the remaining superfamilies to be split into Muricacea and Buccinacea, and it is between these two superfamilies that lines of division may most obviously be drawn. Taylor & Sohl (1962) noted about 800 genera and subgenera in the rachiglossan group. The Buccinacea, with nearly 400, is rivalled for size only by the superfamilies Rissoacea and Cerithiacea amongst all the prosobranchs. A difficulty arises at this point in relation to the number of species which have been described. Many neogastropods are not intertidal in occurrence. Their capture is dependent upon dredging, a method which can often do no more than sample a few isolated spots on the ocean bed. Many species have been described on the basis of these samples without any real knowledge of the variation whjch may affect populations. It seems, indeed, probable that many of these are no more than local varieties, especially when it is remembered that the anatomy of many is very imperfectly known. We have, therefore, been conservative in nomenclature and tended to use broad generic groupings where others might have used narrower ones. The latter may be right, but it is prernature to be sure of this.