REVISION OF ANACYCLUS 103 



according to the principle that the degree of morphological relationship can be equated with the 

 degree of phylogenetic relationship. This, of course is not so, since any concept of overall resem- 

 blance does not have the ability to distinguish between mosaic evolution, convergence and 

 parallelism. The fundamental difference between the methods of the pheneticist and the phylo- 

 genetic systematist is that the latter breaks up the simple concept of resemblance (Hennig, 1965). 



The concept of resemblance can be divided into various categories of (i) resemblance due to 

 convergence, (ii) resemblance due to common possession of primitive (plesiomorphous) character 

 states, and (iii) resemblance due to common possession of derived (apomorphous) character states. 



Convergence occurs because similar organs have evolved adaptations to similar functions from 

 morphological foundations in different organs, so that the character resemblances are merely 

 analogous. Classifications based on resemblances due to convergence then produce polyphyletic 

 rather than monophyletic groups (Fig. 8 C). 



Fig. 8 Group formation and resemblance. Monophyletic groups (A) are recognized by resembl- 

 ance due to synapomorphy (shared derived character states) ; paraphyletic groups (B) occur as a 

 result of symplesiomorphy (shared primitive character states); and polyphyly (C) results from 

 resemblance due to independently derived character states (redrawn from Hennig, 1965). 



Even when problems of convergence can be easily removed, as of course in many cases it can, 

 overall similarity is still not a satisfactory criterion on which to base a phylogenetic classification 

 because it will not produce monophyletic groups. This is due to the fact that characters can 

 remain unchanged through a number of speciation processes. Therefore the common possession 

 of primitive (plesiomorph) characters which have remained unchanged cannot be evidence of the 

 close relationship of their possessors. A classification based on agreement of resemblance due to 

 shared primitive characters thus produces paraphyletic groups (Fig. 8 B). In botanical systematics 

 there has clearly been an obsession with the possession of primitive characters in common, the 

 obvious consequence being that a large majority of angiosperm classifications are paraphylies. 

 It is possible to cite a whole range of examples of obvious paraphylies at various taxonomic levels, 

 e.g. the Ranales, the Bombacaceae, the Heliantheae, Felicia, Aster, Leucanthemum and Chrys- 

 anthemum sensu lato. The supposition that two or more species are more closely related to one 

 another, and that together they form a monophyletic group, can only be confirmed by demon- 

 strating their common possession of derived characters (or synapomorphies). When such 

 characters have been demonstrated, then the supposition has been confirmed that they have been 

 inherited from an ancestral species common to only the species showing these characters. 



Once this premise has been accepted, i.e. that monophyletic groups can be recognized only 

 when morphological resemblance is due to synapomorphy, then the practical aspect can be 

 described. Using Hennig's (1965) Argumentation plan (Fig. 9), every group formation of any rank 

 must be demonstrated by synapomorphous characters. All species or groups of species have a 

 sister group in the modern flora irrespective of divergence through time. Sister groups will thenform 

 monophyletic groups of higher rank. It follows, since evolution occurs through the change of one 

 or more characters, that any one particular character must always occur in a more primitive 

 (relatively plesiomorphous) condition in one group than its sister group. The same is true for the 

 other group with regard to other characters. Therefore, itfollows that there is amosaic distribution, 

 or heterobathmy, of character states in any group and there can be no solely primitive or solely 



