Eight phosphatic porcelain sherds recovered from various historical sites in Charleston were analyzed by electron microprobe. Some sherds contain sulfur (2.3-3.1 wt.% SO 3 ); others contain only traces of this component. The analytical data suggest that the sulfurous sherds are Bow porcelain (London, Bowcock period, ca. 1755-1769). The origin(s) of the low-S samples remains unidentified; one compositionally resembles "gold-anchor period" (phosphatic) Chelsea porcelain (London, ca. 1756-1769) but its decoration is inconsistent with known wares produced by this factory during that era. The degree of vitrification is highly variable, particularly among the SO 3 -poor samples. The melt phase is strongly enriched in incompatible elements (Ti, Fe, Na, K). The phosphate phase [calcined bone ash (hydroxyapatite)] in poorly vitrified samples hosts minute melt blebs, but remains porous. With increasing vitrification, these melt blebs increase in size and begin to coalesce, ultimately forming ameboid patches up to ~10 mm in diameter. In the most vitrified samples, the coalesced melt "leaks" into the matrix, leaving behind a phosphate phase that lacks pores and melt and has a lower CaO/P 2 O 5 ratio (ϭ2.7, molecular proportions) than either hydroxyapatite (3.3) or b-whitlockite (3.0). The two varieties of phosphate occur in some poorly vitrified samples, suggesting the recycling of highfired wasters (as "grog") in their ceramic pastes. Melt compositions vary with contiguous mineralogy, accounting for their divergence from the ternary eutectic in the Ca 3 (PO 4 ) 2 -CaAl 2 Si 2 O 8 -SiO 2 system. The resorption of phosphate by the matrix melt virtually precludes recognition of anorthite formed by the "non-phosphate glass equation."