The classification and interpretation of Amphiboles

Abstract

This thesis reports a study of 2650 published chemical analyses of members of the Amphioole group of minerals with a view to deriving a satisfactory nomenclature for the amphiboles, to outline the overall chemical variation exhibited by the group and the relation between amphibole composition and mode of occurrence. To achieve these' aims a computer program was developed to calculate atomic formulae and simplified or 'basic atomic formulae’ from chemical analyses using an allocation scheme of cations between groups of siter derived from published experimentally determined cation distributions. The concept of the basic atomic formula was also used to make estimates of the iron oxidation state in electron microprobe analyses of amphiboles. A large computer data base using a relational model of data is described and the value of such a data base strategy for the Geological Sciences is briefly discussed. In order to achieve a simple and rational nomenclature an amended and extended version of 'amphibole compositional space' was devised in which calcic (Na (^X) < 0.66, Mg (^X) < 0.66), soda calcic (Na (^X) 0.66 – 1.33, Ng (^X) < 0.66) alkali (Na (^X) > 1.33) amphiboles were named according to the dominant component of their basic atomic formulae. Natural amphiboles close to miyashiroite and sundiusite were located and the validity of both previously discredited names discussed. An extended compositional space was devised to indicate the relation between calcic, soda calcic, alkali and FeMgMn (Mg (^X) > 1.33) amphiboles. From the observed distribution of basic atomic formulae in amphibole compositional space and extended compositional space the importance of coupled Isomorphous substitutions within the amphibole structure and the extent of solvi within the amphibole group was considered. When the mode of occurrence was considered a number of interesting aspects of the complex relationship between amphibole composition and mode of occurrence became apparent.