REVISION OF ANACYCLUS 89 



The latter taxon was considered to be conspecific with A. clavatus Pers. Other known taxa were 

 included in the subtribe VII Artemiseae group 1 Santolineae through the common possession of 

 the following characters 'Capitulis multifloris; corollis staminigeris tubum plano-obcompressum 

 et bialatum stylumque 2 - fidum gerentibus aut teretibus, si achaenium aut cor. $ tubulosa piano 

 obcompressa et bialata est; rhachide bracteolata'. Necker's genus Hiorthia was maintained and 

 taken to comprise H. valentinum ( = Anacyclus valentinus L.), H. aureus ( = Anacyclus aureus L.), 

 H. orientalis ( = A. orientalis) and H. alexandrinus ( = A. alexandrinus Willd.). The most puzzling 

 innovation was the description of a new monotypic genus Cyrtolepis (Linnaea 6 : 166 (1831); Syn. 

 Comp.: 258-259 (1832)) for C. monanthos (L.) Less., based on Santolina terrestris Forsk. ( = Tan- 

 acetum monanthos L.), a taxon clearly conspecific with Anacyclus (Hiorthia) alexandrinus. 



De Candolle (Prodr. 6 : 14-18 (1838)) was considerably influenced by the work of Persoon, 

 Cassini and Lessing when he reclassified the Anthemideae. His main contribution was to use any 

 features which formed natural groups. The homogamous, discoid members having affinities with 

 Anthemis were placed in Cassini's genus Lyonettia particularly to accommodate the anomalous, 

 dwarf, eastern Mediterranean annual Anacyclus cretica L., then considered to be two separate 

 species L. pusilla Cass. and L. rigida DC. Following Cassini's original suggestion, Anacyclus 

 itself was divided into three sections, the first two based on the disc corolla-lobe callosities. 

 Section 1. Pyrethraria DC. accommodated the perennial Anacyclus pyrethrum and section 2. 

 Diorthodon DC. included most of the annual species: A. pulcher Besser ex DC. ( = A. officinarum 

 Hayne), A. tomentosus (L.) DC. and A. clavatus (Desf.) Pers., A. pedunculatus (Desf.) Pers., A 

 radiatus Loisel. and A. valentinus L. The third section, ///0/*//z/0 (Necker) DC., was simply a new rank 

 for Necker's genus to accommodate the poorly understood A. orientalist. (H. orientalis (L.)Necker), 

 now considered to be a discoid member of the Anthemis montana complex (Fernandes, 1975a). 

 De Candolle removed Lessing's anomaly by uniting Tanacetum monanthos L. and Anthemis 

 alexandrina Willd. and called it Cyrtolepis alexandrina (Willd.) DC. 



(4) The period after De Candolle was mainly an exploration phase in which the acquisition of 

 new material, particularly through French, English, Italian and Swedish expeditions to the 

 Maghreb countries and the eastern Mediterranean, resulted in a wealth of notes, minor records 

 and several descriptions of new species. At least eleven new species and many infraspecific taxa 

 were described, particularly by Ball, Maire, Litardiere, Murbeck, Boissier and Reuter, the 

 most recent new species being A. latealatus Huber-Morath. 



By the turn of the twentieth century Anacyclus had become an unnatural group, in an evolu- 

 tionary sense best described as a polyphyletic genus, containing elements of the genus Anthemis 

 (section Hiorthia, Arthrolepis Boiss.) and the Achillea assemblage (Cyrtolepis, Leucocyclus Boiss.), 

 as in the treatment of Bentham (Bentham & Hooker, Gen. PL 2 (1) : 419 (1873)) and Hoffman 

 (in Engler & Prantl, Pflanzenfam. (4)5 : 272 (1894)). As a result of these typo-morphological 

 classifications it has usually been reckoned that the closest affinities of Anacyclus are with 

 Anthemis, itself a polyphyletic taxon. It is now reasonable to hypothesize that the affinities of 

 Anacyclus cretica L. and A. orientalis L. are with Anthemis (Grierson & Yavin, 1975). Most 

 people would agree that Achillea is a very distinct genus clearly separable from the Anacyclus 

 group, but Litardiere & Maire (1924) blurred the distinction when they named a new and unusual 

 alpine species from the Atlas mountains as Anacyclus atlanticus. More recently, Humphries (1977) 

 clarified the status of this species, and Grierson (19756) succinctly compared and contrasted 

 Anacyclus with its sister genus Leucocyclus. This paper now presents a complete revision of 

 Anacyclus and gives an analysis of its phylogenetic relationships. 



Delimitation and systematic position 



The two genera Anacyclus and Leucocyclus are distinguished from related genera by their large 

 anterior-dorsally compressed fruits with lateral wings and continuously thickened pericarp walls 

 (Figs. 4, 5). There is little doubt that the type of fruit compression present in a number of different 

 groups of the Compositae-Anthemideae has evolved several times (i.e. in various S. African 



