The ganophyllite-group minerals, with general formula (K, Na, Ca)xMn6(Si9Al)O24(OH)4·nH2O (ganophyllite = K, eggletonite = Na, tamaite = Ca), are complex modulated layer silicates that contain 2:1 trioctahedral layers with Mn2+-O6 octahedra. Pervasive superstructures have frustrated the numerous attempts at solution of the atomic arrangement of these modulated layer silicates. An orthorhombic dimorph of tamaite has been discovered and its atomic arrangement solved, elucidating the elusive structure of the ganophyllite-group minerals. The dimorph was discovered in the Val Graveglia mining district in the northern Apennines, approximately 35 km east of Genoa, Italy. The phase crystallizes in space group Pnma, a = 16.8146(6), b = 25.2036(9), c = 13.3866(5) Å. The weak reflections from the 3× “superstructure” along a, long observed but never successfully measured in ganophyllite-group minerals, were obtained using a CCD detector and subsequently the atomic arrangement was solved and refined (R = 0.079). The structural modulation in ganophyllite-group minerals results from the misfit between the sheets of Mn2+O6 octahedra and silicate tetrahedra. The atomic arrangement consists of corrugated T-O-T layers, with inverted tetrahedra in the tetrahedral sheets connecting adjacent layers along b. The inverted tetrahedra exist as four-member rings, and incorporate Al, with a maximum Al occupancy of Si2.00Al2.00. Charge balance for the substituent Al is maintained by adding cations (Ca, K, Na) or Al sufficient to balance the charge lost by the Al ↔ Si substitution in the interlayer tetrahedra. Zeolitic H2O molecules also exist in the interlayer channels. Contrary to earlier speculation, the 5.6 Å “subcell” (along a in the orthorhombic dimorph) observed in ganophyllite-group phases is not a true subcell, in that it does not have approximately equivalent atoms at (x, y, z), [~(x + 1/3), ~y, ~z], and [~(x + 2/3), ~y, ~z) for all atoms. Although the majority of the cation scattering (all Mn + 2/3 of the Si atoms) exist in a supercell-subcell relationship, as manifested in very strong h = 3n and very weak (heretofore immeasurable) h ≠ 3n reflections, the lack of such a relationship for all atoms prohibited a successful solution based on previous assumptions of a subcell-supercell relationship.