It was argued that, in contrast to all known modern scleractinian corals that form aragonite skeletons, the original mineralogy of the Cretaceous "Coelosmilia" (ca. 70-65 Ma) was calcite during a period when the Mg2+/Ca2+ ratio of the seawater was presumably the lowest in Phanerozoic (Science 2007 318: 92-94). This is consistent with the idea that scleractinians extend only partial control over their skeletal mineralogy, because calcite is the preferred calcium carbonate polymorph (over aragonite) during abiotic CaCO3 precipitation in low Mg2+/Ca2+ seawater conditions. However, subsequently it was found that a number of different Cretaceous corals formed purely aragonite coralla in the same low Mg2+/Ca2+ conditions, suggesting instead that biological control over the skeletal mineralogy differs between taxonomically different corals (Geology 2017 45:319-322) and posing the question why was "Coelosmilia’s" skeletal mineralogy affected by the seawater chemistry changes and not that of the other corals. An unusual feature of "Coelosmilia" is the infilling of a significant portion of the calice with dense, banded skeletal deposits. They are crystallographically continuous with earlier formed layers of septa, wall and axial structures, but show distinct biogeochemical signatures. To date, no modern analogs of such corals were known, and it was tempting to argue that this skeletal infilling represents secondary, purely abiotic deposits (similar to marine cements, which would be calcitic in Cretaceous seas). Here we document a living scleractinian coral that exhibits a structural organization identical to Cretaceous "Coelosmilia", including dense calicular infilling that is a species-specific feature. Surprisingly, the inner parts of these skeletal elements are composed of fibrous calcite, whereas the rest of the skeleton is aragonitic. We suggest that this modern coral is phylogenetically related to "Coelosmilia" and capable of formation of strictly biologically controlled skeletal mineralogy, which can include both aragonitic and calcitic structures. This original bipartite skeletal organization may explain several diagenetic phenomena that we observe in "Coelosmilia", and is highly relevant for the proper geochemical sampling of corals with such skeletal organization.