Continental flood basalt provinces (cfb) have isotopic ratios different from midocean ridge basaits, characteristics that may be due to (a) interaction of asthenospheric melts with old crust, (b) melting of enriched continental lower lithosphere, (c) upwelling of deep mantle plumes containing recycled components, (d) mixing of enriched and depleted mantle sources, or (e) combinations of these processes. It is apparent that several aspects of the chemistry of cfb are subcrustal in origin and therefore crustal contamination cannot be invoked as a generally applicable hypothesis. In addition wholesale extraction of cfb from the lower lithosphere is unlikely because the lithospheric mantle is believed to be a rather thin (< 150 km) reservoir of cold, anhydrous, granular peridotite. Dry melting of such peridotites would not produce a normal 'basalt' because they have already experienced a melting event. Metasomatism or enrichment processes can enhance the chemical budget of the lower lithosphere thus providing an adequate reservoir for small volume alkaline and potassic melts. However no conclusive evidence exists that such modification of the lower lithosphere is widespread enough to generate a laterally continuous, wet, enriched reservoir. It would appear that cfb require a dominant sub-lithospheric component (plume) to account for the large volumes of magma, in some cases produced over a short time period (e.g. Deccan). If so, the heterogeneity observed in mantle-derived magmas within flood basalt provinces reflects inhomogeneities within plumes rather than relatively shallow inhomogeneities in the lithosphere. The apparent concentration of depleted (STSr/Srsr < 0.7045) cfb in the northern hemisphere and enriched cfb (STSr/arSr ~' 0.7045) in the southern hemisphere may be linked to the preponderance of enriched plumes in the Southern Oceans. Since the early work of Geikie (1987) and Washington (1922), plateau or continental flood basalts (cfb) have been thought of as sub-crustal in origin because of their size and apparent relationship to dike injection within the crust. The first attempts to define the origin of cfb were by Thompson et al. (1972) and McDougall (1976) who believed that cfb originated in the mantle. Thompson (1977) deduced that certain cfb were extruded at rates equivalent to mid-ocean ridge basalts. To account for these large volumes, Thompson (1977) and Swanson et al. (1975) proposed that cfb had an origin similar to oceanic basalts and that any modification of their trace element or isotopic ratio was due to digestion of variable quantities of crust during upwelling or storage of the magmas in the crust. Whilst this is undoubtedly the case for several cfb, including the British Tertiary, many others retain evidence of mantle heterogeneity. A classic example is the Snake River Plain, Idaho, where Leeman (1975) reported enriched tholeiities with high 87Sr/86Sr ratios that lacked unequivocal evidence for wholesale crustal contamination. Consequently, Leeman (1975) proposed that these enriched tholeiitic melts were derived from the continental keel or sub-continental mantle, a proposition supported by Brooks & Hart (1978) who argued that flood basalts inherited age information from their lithospheric mantle source region. This suggestion of a shallow source for cfb contrasted with that proposed by De Paolo & Wasserburg (1979) and Wasserburg & DePaolo (1979). According to their neodymium isotopic data cfb originated in an undifferentiated (chondritic) source similar to the source of ocean island basalts (OIB) in the lower mantle. Despite these sub-lithospheric models for the genesis of cfb, evidence from mantle xenoliths helped galvenize models involving enriched continental lithosphere (Hawkesworth et al. 1983; Menzies et al. 1983; Carlson 1984; McDougall 1988; Ellam et al. 1991). Isotopic and elemental data on basaltborne and kimberlite-borne xenoliths led to a widespread acceptance that much of the lower lithosphere was chemically enriched and as such a suitable reservoir for cfb (Basu & Tatsumoto 1980; Menzies & Murthy 1980; Menzies & Hawkesworth 1987; Nixon 1987). Pioneering work From STOREY, B. C., ALABASTER, T. & PANKHURST, R. J. (eds), 1992, Magmatism and the Causes of Continental Break-up, Geological Society Special Publication No. 68, pp. 31-39. 31 2013 at Royal Holloway University of London on July 10, http://sp.lyellcollection.org/ Downloaded from