Bruyns (1999) has responded negatively to my paper (Liede, 1997) that suggested a classification for the large tribe Asclepiadeae (264 names for 100 genera). It is easy to find fault with any large-scale classification based on morphological characters but not so easy to suggest a better one. So it is not surprising that Bruyns picks at the exceptions to the synapomorphies outlined (he seems to eliminate the possibility of reversals of single characters completely) without suggesting a better placement for even a single genus. A tribe, a monophyletic unit (its monophyly is not disputed even by Bruyns), has been analysed. Bruyns argues several times that a character cannot serve as a synapomorphy within the tribe if it has been found in a few members of other, only distantly related, tribes. The sister-group of the Asclepiadeae the only group that matters for the decision as to whether or not a character is synapomorphic is, to the best of present knowledge, the tribe Marsdenieae (Liede, 1996b, Sennblad & Bremer in Sennblad, 1997; Civeyrel & al., 1998), so I will not consider further arguments involving members of the tribes Secamoneae or Stapelieae. Also, characters that occur in single, isolated species of large genera, such as the elevation of the gynostegium and corona on a column in Dischidia bengalensis Colebr. and Hoya lanceolata D. Don, have to be considered as independently derived in these genera and cannot be used for higher-level comparisons. It seems that for the sake of his argument Bruyns is excluding the existence of parallelism even though he should know of its high incidence in Asclepiadaceae. Bruyns also seems unaware of the principle of priority in nomenclature, else he would know that anyone employing a subtribal concept cannot avoid using the names adopted by previous workers, whenever the group concerned includes the genus upon whose name the name of the subtribe is based. Whereas I am indeed adopting a subtribal classification in such a large tribe as the Asclepiadeae, the vastly different circumscription of almost all of my subtribes from those of Schumann (1895) renders his accusation of my rehashing an old concept untenable. In the Asclepiadinae, the close relationship of a genus (Asclepias) with tuberous roots I am indeed grateful to Bruyns for his careful analysis of this characterand one without (Calotropis) has been confirmed by molecular analysis (Civeyrel & al., 1998; Sennblad, 1997), and I have resolved the position of the tuberless and, in some species, twining genus Pergularia (Liede, 1996a). Whereas decumbent or spreading is a growth form easily derived from an erect one, not even Bruyns claims to know another twining member of the Asclepiadinae. The position of the Gonolobinae as a subtribe of the Asclepiadeae, suggested by Swarupanandan & al. (1996) and formalised by Liede (1997), has since been confirmed by two independent molecular analyses (Sennblad, 1997; Civeyrel & al., 1998). The same analyses suggest a sister-group relationship between all the genera analysed that belong to the Oxypetalinae (sensu Liede, 1997) and those that belong to the Gonolobinae (sensu Liede, 1997). In the paper of Swarupanandan & al. (1996: