In the past century Cornaceae were mostly delimited in a wide sense and they represented a fairly heterogeneous assemblage. HARMS (Ber. Deut. Bot. Ges. 15, 1897, 28 and in E. & P. Nat. Pfl. Fam. 3, 8, 1898, 255) distinguished 7 subfamilies. Of these Garryoideae were later mostly recognized as a separate family Garryaceae, as Alangioideae Alangiaceae, Nyssoideae and Davidioideae together as Nyssaceae, leaving Cornaceae with the remaining three subfamilies Cornoideae, Curtisioideae (monotypic, South Africa) and Mastixioideae (monotypic, Indo-Malesian tropics). Cf. WANGERIN, Pfl. Reich Heft 414 (1910) 18. In recent years, however, the other genera (6) of the Cornoideae, besides Cornus, have also been recognized as monotypic families, with the exception of Corokia which was transferred to Saxifragaceae-Escallonioideae. Notably TAKHTAJAN (Proiskh. Prokruitosem. Rast.: 89, non vidi) is in favour of these monotypic families. In his ‘Flowering Plants’ (ed. C. JEFFREY; 1969: 227) he accepted 7 segregate families besides Cornaceae sens. str. (omitting mention of two Madagascan genera, one of which he had formerly also raised to family rank, according to SHAW, 1973). These 7 families he arranged, together with Araliaceae and Umbelliferae, in the order Cornales, a phylogenetic construction of affinity not much different from earlier conceptions. The general impression is thus that the distinction of the segregate families is largely an inflation in rank.