Notes: FNA8: "As typically described (e.g., A. Cronquist 1981; V. H. Heywood 1978), Primulaceae were clearly polyphyletic, closely related to Myrsinaceae and Theophrastaceae. M. Källersjö et al. (2000) and B. Ståhl and A. A. Anderberg (2004) removed the nonrosette terrestrial members from Primulaceae in the broad sense and placed them in the Myrsinaceae, which are further distinguished by leaves and calyx often dotted with yellow or dark streaks, flowers with relatively shorter corolla tubes, seeds immersed in placentae, and wood devoid of rays or with multiseriate rays only. Maesa, consisting entirely of trees found in the Eastern Hemisphere tropics, also has semi-inferior ovaries, pedicels with two bracts, and wood with both uniseriate and multiseriate rays; it, too, was removed from Primulaceae/Myrsinaceae and placed in its own family (Källersjö et al.). The families Primulaceae in the narrow sense, Myrsinaceae, Theophrastaceae (including Samolaceae), and Maesaceae then form a monophyletic clade within Ericales (P. F. Stevens, http://www.mobot.org/MOBOT/research/APweb/), sharing some features, most notably flowers with sympetalous corollas, stamens in a single series and opposite the petals, free-central placentation, bitegmic, tenuinucellate ovules, and plants generally with tannins and saponins.

Additional evidence (L. Martins et al. 2003) indicates that Androsace and Primula may not be monophyletic; more work is needed to resolve these issues. The work of M. Källersjö et al. (2000) showed that Douglasia should remain separate from Androsace, and Dodecatheon should remain separate from Primula, although Dodecatheon clearly is derived from Primula subg. Auriculastrum. Alternative views suggesting more inclusive concepts of Primula and Androsace have been offered by I. Trift et al. (2002), A. R. Mast et al. (2004), and G. M. Schneeweiss et al. (2004). The phylogenetic position of Cyclamen, a scapose taxon currently included in Myrsinaceae, has not been resolved. Our understanding of Primulaceae is still in flux, and future taxonomic realignments at the familial and generic levels are to be expected."

Family information last updated 10/29/2009 by David Giblin.