Controversy concerning the age of regional granulite-grade metamorphis
m in Sri Lanka stems from the fact that different isotopic systems app
arently record different events. Published isotopic data show that zir
cons from igneous and sedimentary protoliths experienced considerable
Pb-loss at 550 Ma and that new growth of zircon, monazite, rutile and
garnet between 539 +/- 6 and 608 +/- 3 Ma probably dates near-peak met
amorphic conditions. Upper concordia intercept ages between approximat
ely 70 and approximately 1950 Ma for magmatic zircons from high-grade
orthogneisses and charnockites date igneous emplacement and also suppo
rt the occurrence of granulite metamorphism after approximately 670 Ma
. A granulite of igneous origin from the Kataragama klippe in southern
Sri Lanka yielded discordant ion microprobe data that suggest emplace
ment of the protolith at 1770-1880 Ma. Thus a Rb-Sr whole-rock age of
1930 +/- 130 Ma from the same suite is unlikely to reflect isotopic ho
mogenization during granulite formation, and we interpret this and oth
er best-fit lines as a memory effect in the Rb-Sr isotopic system that
provides approximate ages for the gneiss protoliths, not for the high
-grade metamorphism.