Abstract: Our knowledge of the nature, generation and maintenance of largescale biodiversity patterns is still far from complete. This is particularly so in the Southern Hemisphere and in the marine realm, where recent taxonomic investigations of Mollusca and other invertebrate groups has cast doubt upon the existence of a simple cline in species richness between the tropics and the pole. Comparatively high regional diversity values for the shelled gastropods and other epifaunal taxa implies a considerable evolutionary legacy; this is supported, at least in part, by available evidence from the fossil record. Certain families within the living gastropod fauna maintain their prominence when traced back 40 m.y., and perhaps even longer; in addition, several Southern Ocean gastropod and bivalve genera can now be traced back to at least the late Eocene. Use of a variety of refugia may have enabled many taxa to survive repeated glacial advances. As we begin to revise our concept of the nature of latitudinal diversity gradients, so we also need to examine regional variations in evolutionary rates. Clearly this is a complex issue. but recourse to a pilot study based on the molluscan fossil record suggests that there may be no significant difference between the rates of radiation of tropical and cold-temperatdpolar taxa. The most diverse clades, which are all tropical, are simply the oldest. What data are available from the fossil record indicate that there is no appreciable latitudinal variation in rates of extinction either. Time, but not necessarily environmental stability, would appear to be crucial to the development of pockets of high taxonomic diversity. Recent improvement in our understanding of the biology of many polar marine invertebrates suggests that life in cold water is not an insuperable evolutionary problem. Of qual importance to any intrinsic properties of organisms which may have governed the differentiation of large-scale biodiversity patterns is the role of extrinsic processes. Foremost among these has almost certainly been repeated range shifts in response to Cenozoic climatic cycles.