Re-examination of the mercury-rich regions of the Ca-Hg and Sr-Hg phase diagrams has shown that the phases previously identified as "AHg(3.6)" should be reformulated as A(11-x)Hg(54+x) (A = Ca, Sr). The crystal structures for representative members of these A(11-x)Hg(54+x) phases were determined from single-crystal X-ray diffraction data (Pearson symbol hP65, space group P (6) over bar; a = 13.389(1) angstrom, c = 9.615(1) angstrom for Ca10.92(2)Hg54.08 (x = 0.08(2)); a = 13.602(2) angstrom, c = 9.818(1) angstrom for Sr10.48(4)Hg54.52 (x = 0.52(4))) and confirmed by powder Rietveld refinements (R-B = 0.020 for Ca10.7(2)Hg54.3 and 0.014 for Sr10.7(3)Hg54.3). Diverse coordination polyhedra surround the A (CN14-16, multiply capped pentagonal or hexagonal prisms as well as Friauf polyhedra) and Hg atoms (CN11-13, pentacapped trigonal prisms and icosahedra). Partial disorder of Hg into one of the A sites accounts for the nonstoichiometry in the A(11-x)Hg(54+x) phases. If this disordered A site is completely occupied by Hg atoms, the composition is constrained to a maximum of x = 2 in A(11-x)Hg(54+x), corresponding to a small homogeneity range of "A(0.14-0).Hg-17(0.86-0.83)"; the true homogeneity range is likely narrower. The structure can be regarded as being built up from a stacking of triangular nets with hexagonal voids that are filled with single atoms or various clusters. In particular, the presence of triangular Hg-3 clusters in ordered orientations distinguishes this structure from that of the related Gd14Ag51-type structure, in which triangular Ag-3 clusters are in disordered orientations. Band structure calculations reveal a small degree of electron transfer from the A to Hg atoms, supporting the presence of a partially anionic mercuride substructure.