Tamazert is an elongate complex 18 km in length and up to 6 km wide which has been emplaced along a major thrust plane in the core of a large anticline on the northern margin of the High Atlas, the intrusion varying between 1800 and 2970 m above sea level. It cuts Liassic to Cretaceous limestones and marly limestones which lie within a graben (Agard, 1960). Four distinct phases of igneous activity can be distinguished. The earliest phase is represented by two bodies of micaceous pyroxenite which lie towards the centre of the complex. These were followed by nepheline syenite which is by far the most abundant rock type and which takes a somewhat flattened annular form. The third phase consists of carbonatites which locally are associated with small amounts of trachybasalt, nephelinite and analcimite. The final phase consists of numerous dykes of a wide variety of petrographic types which cut all the earlier rocks (Jérémine, 1949; Agard, 1960; Bouabdli et al., 1988). The pyroxenites are variable consisting of differing proportions of biotite, barkevikitic amphibole, augite and aegirine-augite, olivine, titanite, perovskite, apatite and titanomagnetite. The nepheline syenites are heterogeneous rocks a range of different types of which have been described by Jérémine (1949); they include inclusions of pyroxenite and marble. They comprise orthoclase, sodic plagioclase, nepheline, cancrinite, sodalite, hauyne, analcime, biotite, hornblende, arfvedsonite, aegirine-augite and aegirine, eudialyte and Ti-rich garnet (Jérémine and Dubar, 1947; Agard, 1960). Nepheline syenite pegmatites are common and comprise orthoclase, albite and some perthite, aegirine, nepheline, fluorite, cancrinite, sodic amphibole, sodalite, zircon, calcite, galena, pyrite, rinkite, lavenite, astrophyllite and eudialyte. Xenoliths of limestone within the nepheline syenites are metamorphosed often having marginal concentrations of garnet, vesuvianite, diopside and amphiboles. One group of syenites, occurring particularly in the northern part of the complex, are described by Agard (1960) and others as 'metasyenites', they are probably fenites (Aghchmi, 1984) which have been generated by emplacement of the carbonatites, but other rock types may also have caused fenitization. Many of them are characterised by high potassium values and they have generally been formed at the expense of the nepheline syenites. They consist essentially of orthoclase, sometimes perthitic, muscovite, iron oxides, pyrite, apatite and zircon. The carbonatites form two principal bodies which are irregular in outline and about 1 km in length, and numerous dykes and small, irregularly shaped intrusions. A detailed map of an area in the southwest of the complex (Aghchmi, 1984) demonstates the abundance of carbonatite intrusions and their concentration within 'metasyenites'. The carbonatites are variably fine- or coarse-grained or pegmatitic and may be massive or form bodies, sometimes pipe-like, of breccia. The carbonate is either calcite or ankerite; apatite is ubiquitous and second in abundance to carbonate. Pyrochlore is plentiful and other minerals identified include albite, orthoclase, pyrite, magnetite, barite, synchysite, parisite, monazite, strontianite, celestite, blende, galena and fluorite. Descriptions of these minerals together with some analytical data are given by Aghchmi (1984) who also presents rock analyses, including trace element data, of four facies of carbonatite. Mourtada (1997; Mourtada et al., 1997) has identified extrusive as well as intrusive carbonatites which are dolomitic. They are associated with diatremes in the Issli and Tisslit areas and are estimated to comprise about 10% of the Tamazert carbonatites. The extrusive carbonatite forms spherical lapilli comprising some 30% of layered tuffisites and breccias and consists of dolomite and/or ankerite, orthoclase, a little calcite, rutile and Cr-spinel; analyses of the mineral phases and of separated lapilli are given. The dykes comprising the final stage of igneous activity at Tamazert include carbonatite, phonolite, tinguaite, syenite, pegmatites and alkaline lamprophyres (Bouabdli et al., 1988). Field relationships suggest that all dyke types were emplaced at about the same time; the lamprophyres are the most abundant and have been investigated in some detail by Bouabdli et al. (1988). The lamprophyres are radially distributed around the carbonatites and form a swarm cutting limestones to the southeast of the complex. Individual dykes are up to several kilometrtes in length and 3 m thick; they are monchiquites. They contain phenocrysts of clinopyroxene (aegirine cores to diopsidic rims), mica (Ti-rich biotite cores and phlogopite rims), kaersutite, olivine (Fo89-83) and rare apatite in a groundmass of pyroxene, amphibole, phlogopite, alkali feldspar, and analcime or nepheline; perovskite and melilite are scarce. Ocelli are characteristic and consist of either carbonate or K-feldspar, zeolite and subordinate pyroxene and biotite. Analyses of pyroxene, amphibole and mica are given by Bouabdli et al. (1988) who also include numerous analyses of lamprophyre, nepheline syenite, phonolite, tinguaite and carbonatite; much trace element data are included. A detailed study of C, O, Sr and Nd isotopes in dykes of lamprophyre, syenite, phonolite, carbonatite and breccia has been made by Bernard-Griffiths et al. (1991). Salvi et al. (2000) have made a study of fluid inclusions in nepheline syenites and pegmatites aimed at understanding hydrothermal mineralisation of high field strength elements.