Abstract
A re-evaluation of AlIVvs. AlVI occupancy of calcic and subcalcic amphiboles of the literature recalculated to 23(O) suggests that several analyses from specimen 102–1 are more tschermakitic than any amphiboles previously reported regardless of the Fe2O3 approximation used. A standard A1VIvs. A1IV plot (such as that described by Leake in 1971) is unsatisfactory because it fails to distinguish the A1VI content due to glaucophane-type [Na(M4),AlVI → Ca,Mg] substitutions and the A1IV content due to edenite-type substitutions [Na(A),AlIV → □,Si].
On an [AlVI − Na(M4)] vs. [AlIV + Na(M4)] plot, the alkali-rich end-members of hornblende are shifted away from tschermakite. Such a plot clearly shows the unusual composition of the southwestern Massachusetts amphiboles and of natural tschermakitic amphiboles in general. The approximate substitutional limitation AlVI − Na(M4) = 0.75[A1IV + Na(M4)] changes slightly to 0.82[A1IV + Na(M4)] when Fe3+, Ti4+, and Cr3+ are included with A1VI, suggesting a very strongly defined limitation of maximum octahedral R3+ occupancy in calcic amphiboles.