Continental basaltic magmatism reveals a compositional diversity that is greater than in the oceanic realm. This relates to the interaction of mafic magmas with the more easily fusible parts of the continental crust, but much compositional diversity has been attributed to magma sources located in the continental mantle lithosphere. Basalts from numerous CFB provinces and extensional continental regimes reveal compositional trends that are consistent with a pattern of early lithospheric melts followed by plume-related or asthenospheric magmas once the lithosphere has thinned (Perry et al., 1987, Fitton et al., 1991, Turner et al., 1996). Heating of the lithosphere may result in so-called thermal erosion during which lithospheric mantle is incorporated into the underlying asthenosphere or mantle plume where it may contribute to plume-related magmatism (Davies, 1994). Thermal erosion melting implies a dynamic system in which melt is extracted by compaction of a convecting matrix, similar to the predominant process operating beneath mid-ocean ridges (e.g. Elliott, 1997, Asmerom et al., 2000). Within the Kenya rift, mafic melts of variable composition have been generated over the past 35 Ma and reflect the interaction between the underlying mantle plume with the overlying lithosphere (Macdonald et al., 2001).

New mass spectrometric analyses of trace element abundances and (²³⁰Th/²³⁸U) disequilibrium from recent (<10 ka) basalts from the Kenya rift have allowed us to confirm dynamic melting processes in the source regions of basalts with lithospheric origins (Rogers et al., 2004). Northern Kenya basalts define a vertical array on the equiline diagram, with all but one of the samples lying between (²³⁸U/²³²Th) of 0.655 – 0.735. This array is consistent with a dynamic melting regime (Elliott, 1997). It is suggested that basalts from the northern Kenya rift represent melts derived from lithospheric material that has been thermally reactivated by the underlying mantle plume and is now behaving as convecting mantle. The minimum upwelling rates beneath the rift axis are ~1 cm yr-1 but basalts with (²³⁰Th/²³⁸U) < 1.4 may be generated either where the upwelling rate is greater or from a source with < 6% modal garnet.

A major challenge in igneous petrology is to determine the rates at which magma evolves, the factors that contribute to controlling those rates and how they relate to the thermal budgets of volcanic systems. An understanding of the evolution of magmatic systems and of the dynamics of the deeper regions of volcanoes is critical to the development of models for accurate forecasting of volcanic hazards (Sparks, 2003).

Despite the high silica contents of the Longonot trachytes, fractionation rates (0.5–2×10^-4/year) are comparable with those from the basaltic system at Ardoukoba, (3.3–3.6×10^-4/ year, Vigier et al., 1999), but the rates may be >2×10^-4/year within the trachytes themselves. A possibly slower rate of fractionation is inferred for the differentiation of hawaiite to trachyte and may reflect a two-stage process in the evolution of the Longonot lavas. Initially, hawaiitic magma differentiates in a large magma reservoir at mid-crustal levels to produce a parental trachyte magma, which is then introduced into the magma reservoir in the immediate sub-volcanic environment, in which the parental trachyte further differentiates into the range of compositions erupted as lavas at the surface.
(²³⁰Th/²³⁸U) disequilibria indicate that limited U/Th fractionation occurred during the past 10 kyr, whereas (²²⁶Ra/²³⁰Th) disequilibria reflect the effect of alkali feldspar fractionation. (²²⁶Ra/²³⁰Th) disequilibria in the trachyte lavas are interpreted using a model that combines the equations of radioactive decay and in-growth with Rayleigh crystallization to give fractionation rates of about 0.2×10^-4/year for the evolution of hawaiite to trachyte, but more rapid rates of up to 3×10^-4/year for fractionation within the trachyte sequence. (²²⁶Ra/²³⁰Th) from two whole-rock–alkali feldspar pairs are interpreted to show the crystals formed at 5800 y and 2800 y , implying that phenocryst formation continued almost up to the time of eruption. The results strongly indicate that fractionated magmas can be stored for periods on the order of 1000–2500 years prior to eruption, whereas other magmas were erupted as fractionation was proceeding.

Recent and historic lavas from Tenerife, Canaries, show that magma differentiation has happened in two stages ( Hawkesworth et al, 2001 Thomas et al. under review). Stage one occurred at a depth of greater than 20 km , where 50% fractional crystallisation produced phonolite from primitive basanite taking in the order of 200 ky in a closed system. The second stage occurred at shallow levels at sea level and phonolite was differentiated and mixed with crustal phonolite to form a more evolved amphibole bearing phonolite by 50% fractional crystallisation. This second stage occurred some few hundred years prior to eruption.
Along destructive plate margins the rates of fluid-induced melt generation may be significantly faster than that beneath ocean ridges and ocean islands (Turner et al., 1996,1997), and it has been suggested that that may in turn be responsible for some of the distinctive trace element features of island arc rocks. Hobden et al., (2002), present geochemical data on the frequently active composite cone of Ngauruhoe, in New Zealand based on the same suite of rocks. Preliminary conclusions are that specific batches of melt evolve within a number of chambers in the system, and are tapped periodically. Each batch provides a different composition with glass matrix and feldspar separate yielding age constraints on the evolution within the chamber.
A number of studies have highlighted the different information available from U-Th-Ra analyses of separated crystals and the groundmass, or of bulk rocks, in studies of the time scales of crystallisation and of magma differentiation (e.g. Hughes and Hawkesworth, 2000). The presence of crystals with ages significantly older than the age of eruption arguably provides maximum ages for the magmatic system (Halliday et al., 1989; Condomines et al., 1995). However, in many cases the separated crystals are not in bulk equilibrium with the host rock, and so the links between the ages of crystals and the time of magma differentiation are difficult to establish (Hawkesworth et al., 2000). Nevertheless, the observation that separated minerals and groundmass from five samples from Soufrière, St. Vincent, scatter around a 50 ky isochron, has highlighted the need to develop better constrained models for magma evolution over longer time scales (Heath et al., 1998; Hawkesworth et al., 2000). Recent and historic Montserrat volcanics yielded data for glass, phenocryst and whole rocks close to secular equilibrium in U-Th- Ra systematics. Such results suggest that initial U excesses due to the slab fluids contribution in the subduction zone have decayed. The fluid signature is less that >350kyr old and this is consistent with the highly differentiated nature of these rocks, which have undergone long term magma storage and evolution in the crust (Zellmer et al 2002).

Halliday et al., 1989, Earth and Planetary Science Letters, 94, 274-290.
Hawkesworth et al., 2000, State of the Arc, abstract.
Hawkesworth et al., 2001, EOS Transactions 82, 261-265.
Heath et al., 1998, Earth and Planetary Science Letters, 160, 49-63.
Hobden et al., 2002, Bull Volcanol 64, 392-409.
Hughes and Hawkesworth, 2000, State of the Arc, abstract.
Thomas et al., 2006. Earth Planet. Sci. Lett.( under review).
Turner et al., 1996, Earth and Planetary Science Letters, 142, 191-208.
Turner et al., 1997, Chem. Geol. Spec. Vol., 139, 145-164.
Zellmer et al., 2000, Earth and Planetary Science Letters, 174, 265-28.