The possibility of an interpretation of the floral region based on the anthocorm concept is demonstrated. By assuming that the reproductive region of the primitive Piperales, Eupteleaceae, and Cercidiphyllaceae (and of some of the principal groups of the Polycarpicae) represent surviving stages of ancient anthocormoid structures or parts thereof, the change-over from an anthocorm to a ‘flower’ is shown to be the result of a number of independent and parallel evolutionary tendencies acting upon one of the three primary types of anthocormoid structures viz. unisexual ones, ambisexual ones with separate, coaxial androclads and gynoclads, and ambisexual ones with gonocladial androgyny. The two ambisexual types originated in connection with the partial switch-over from the gymnospermous condition characterised by dicliny and anemophily to forms of monocliny associated with entomophily in pre- and protoangiospermous groups. A true ‘flower’ is always compounded of a number of coaxial gonoclads, with few exceptions variously associated with other gonoclads of the same kind, with their subtending stegophylls, and/or with the supporting anthocorm axis (= true floral axis). Evidence from various sources is not at variance with this interpretation which implies, among other things, that in numerous taxa of the Polycarpicae the pistils or the constituents of the gynoeciurn traditionally called ‘carpels’ are derivatives of ovuliferous pteridospermous cupules. In the second part of this paper the taxonomic relationships and the floral morphology of the principal ranalean groups and of several other taxa are discussed on the basis of the anthocorm theory. In the third part a brief survey is given of the consequences of the new interpretation for the phyllo-and stachyotaxis in the reproductive region, of its connection with the ontogenetic development of floral parts, and of its meaning for the comparative topology of fertile floral organs.