A brief review of Becke's (1908) replacement and Schwantke's (1909) exsolution models for myrmekite genesis and further consideration of the morphology and spatial distribution of myrmekite as outlined by Phillips (1974) lead to the conclusion that both hypotheses have a place in explaining the origin of myrmekite. The Schwantke model is generally best applied, for example, to high-level undeformed massive granitoids and the Becke model to deformed metamorphic rocks. The latter hypothesis and a third model involving interaction between exsolution and replacement (Ashworth, 1972) may be used especially to explain the association of muscovite and myrmekite in rocks whose mineral assemblages have undergone retrograde modification, involving either redistribution of Na, Ca and K by local metasomatic reactions among the mineral grains, or metasomatic change of bulk rock composition.