Abstract: Analyses of tourmaline and coexisting muscovite in tourmaline-bearing rocks from SE Ireland indicate that the content of Si in tourmaline is not fixed but is dependent on bulk chemistry, varying inversely with Al. Zoning patterns within tourmaline reflect this and, despite theoretical objections, substitution of Si by Al in the tetrahedral site must occur in nature. Ti may also substitute for Si. Ca and Ti are generally involved in a coupled substitution. Estimates of the degree of alkali-defect and proton-loss substitution in SE Ireland tourmalines indicate that the former process is at least as important as the latter and that bulk chemistry is important in determining which substitution occurs. The general importance of alkali-defect substitution and Al⇌Si suggests that the most useful way to treat microprobe data for schorl-dravite tourmalines is to assume the presence of 3B and 4(OH) and calculate on the basis of 24.5 oxygen atoms.