The classifi cation of granitic pegmatites was frequently attempted during the past century, with variable degrees of success and applicability. Internal structure, paragenetic relationships, bulk chemical composition, petrogenetic aspects, nature of parent medium, and geochemical features were applied. However, all schemes were marked by contemporary degrees of understanding of these parameters, and most attempts were hindered by ignoring differences in geological environment. Substantial progress was achieved only since the late 1970s. The classifi cation is approached here from two directions, based on but broadened and refi ned from earlier works by Ginsburg and Čern´y. The fi rst concept deals with geological location, leading to division of granitic pegmatites into fi ve classes (abyssal, muscovite, muscovite -rare-element, rare-element, and miarolitic), most of which are subdivided into subclasses with fundamentally different geochemical (and in part geological) characteristics. Further subdivision of most subclasses into types and subtypes follows more subtle differences in geochemical signatures or P-T conditions of solidifi cation, expressed in variable assemblages of accessory minerals. The second approach is petrogenetic, developed for pegmatites derived by igneous differentiation from plutonic parents. Three families are distinguished: an NYF family with progressive accumulation of Nb, Y and F (besides Be, REE, Sc, Ti, Zr, Th and U), fractionated from subaluminous to metaluminous A-and I-type granites that can be generated by a variety of processes involving depleted crust or mantle contributions; a peraluminous LCT family marked by prominent accumulation of Li, Cs and Ta (besides Rb, Be, Sn, B, P and F), derived mainly from S-type granites, less commonly from I-type granites, and a mixed NYF + LCT family of diverse origins, such as contamination of NYF plutons by digestion of undepleted supracrustal rocks.Keywords: classifi cation, granitic pegmatites, geochemistry, mineral assemblage, petrogenesis. SOMMAIREIl y a eu plusieurs tentatives de classifi cation de pegmatites granitiques au cours du siècle dernier, avec un taux de réussite et une applicabilité variables. La structure interne, les relations paragénétiques, la composition chimique globale, les aspects pétrogénétiques, la nature du milieu de croissance, et les caractéristiques géochimiques ont tous été utilisés comme bases de classifi cation. Toutefois, ces schémas ont été limités par le niveau de compréhension de ces paramètres lors de leur application, et par négligeance des différences du milieu géologique. Des progrès substantiels ont seulement été atteints depuis la fi n des années 1970. La classifi cation est abordée ici de deux directions, fondées sur les travaux antérieurs de Ginsburg et Černý, mais affi nés et considérés dans un contexte élargi. Le premier concept porte sur la situation géologique, et mène à cinq classes de pegmatites granitiques: abyssale, à muscovite, à muscovite -éléments rares, à éléments rares et miaro...

A systematic review of the oxide minerals of niobium and tantalum is presented, including varietal names, crystal chemistry, structural features, relationships to other phases and paragenesis of each mineral group, series or species. A separate section deals with oxide minerals that occur exclusively as secondary, low-temperature phases of largely metasomatic origin. Oxide minerals of Sn, Ti and W that carry substantial Nb and Ta are discussed as well. A new crystal chemical classification ofthe oxide minerals ofNb and Ta is presented.Abriefreview is presented of the crystal chemistry ofthe silicates ofNb and Ta, and ofthe silicates ofTi and Zr that exhibit Substitution ofthese elements by Nb and Ta. Paragenetic affiliation of these silicates is reviewed. Miscellaneous minerals of Nb and Ta (borates, phosphates, borosilicates) arealso mentioned.Most of the oxide minerals of Nb and Ta typically occur in granitic rocks of orogenic affiliation, particularly in rare-element granitic pegmatites. However, some ofthese arealso found in, and a few are characteristic of, anorogenic alkaline igneous sequences from carbonatites and gabbros to nepheline syenites and alkaline granites. In contrast, all silicates ofNb (±Ta)-and (Nb,Ta)-bearing titanosilicates and zirconosilicates are related to the anorogenic parageneses.At present, the oxide minerals are the only important sources of Nb and Ta. Niobosilicates, and mainly some of the titanosilicates and zirconosilicates with significant Nb substitution, represent potential ore minerals forthe future. However, no alternate ore minerals seem to be available for extraction of tantalum.Note added in press: Continuing examination ofTa-rich wodginites has revealed the presence of significant quantities of Li, suggestive of a Li4 Ta120 32 component known as a synthetic wodginite-type phase (Gatehouse BM, Leverett P 1972, Lithium triniobate (V), LiNb30 8 ; Cryst Struct Comm 1 :83-86). This development will require modification ofthe wodginite systematics presented here.

No abstract

"Allanite" is a poorly defined collection of species belonging to the epidote group. In the past, "allanite" was defined as merely being lanthanon-bearing, but more recently, it is defined as being Ln-dominant at the A2 site. Species of the allanite subgroup include allanite-(Ce), allanite-(La), allanite-(Y), androsite-(La), dissakisite-(Ce), dollaseite-(Ce) and khristovite-(Ce). Lack of recognition of the recommendations of Nickel & Mandarino (1987), and of the relation between members of the subgroup with "allanite" in their species name and those without, has led to a number of errors in naming species and in use of the term "allanite". Ten methods have recently been published describing how the relevant formulae may be calculated. Each is reviewed in conjunction with new observations on the behavior of Si, Cr, V, Mn 2+ and A-site vacancies. This evaluation results in a recommended procedure for the calculation of the formula involving a basis of 6 (M + T) cations and 12(O,F,Cl) + 1(OH). Recalculation of the formulae of published compositions of allanite-subgroup minerals shows that some apparently new species are better interpreted as intermediate solid-solutions between conventional end-members; there are, however, at least five potentially new species awaiting description.

No abstract