88 C. J. HUMPHRIES 



idea of re-examining the generic limits within this group emerged. He questioned the division 

 based on the presence or absence of ray florets: 'Genus bifrons; nam accedente radio Anthemis 

 est, sicut Anthemis, radio deficiente, Anacyclus; unde Anacyclus-Valentinus, Anthemis- Valentina, 

 Anacyclus aureus, Anthemis-aurea' ; but in the actual species descriptions he did a rather illogical 

 thing: Anacyclus was considered to be a monotypic genus based on A. aureus, and the remaining 

 taxa were placed in Anthemis. The use of the name Anacyclus aureus L. (Mantissa 2 : 287) appears 

 to be based on a misidentification on Brotero's part, as De Candolle pointed out (Flore Francaise 

 5 : 480 (1815)), since he meant this species to be conspecific with Anacyclus valentinus L. Indeed, 

 today it is known as Chamomilla aurea (Loefl.) Gay ex Cosson & Kralik. 



Persoon (Syn. PL : 464 (1807)), adhering to the concise style typical of his diagnostic conspectus, 

 distinguished between Anthemis and Anacyclus very precisely, in that the former genus has 

 tetragonous or cylindrical cypselas without 'borders', whereas the fruits of the true species of 

 Anacyclus are invested by a lateral membrane. This division is based on only some of the characters 

 which define Anacyclus and so, although it reassociated some of the misplaced Anacyclus species, 

 it had the curious effect of excluding the perennial species Anthemis pyrethrum and retaining the 

 two Mediterranean species Anacyclus creticus and A. orientalis. De Candolle (1815) followed 

 Persoon in his use of the winged fruit as the primary diagnostic feature for the genus, but is more 

 consistent in its application. Consequently, Anacyclus was taken to comprise the five species 

 A. valentinus, A. radiatus, A. purpurascens, A. tomentosus and A. clavatus. De Candolle remarked 

 that 'Ces cinq plantes ne forment peut-etre qu'une seule espece. Comme je n'ose cependant 

 raffirmer absolument, je vaisindiquer ici un peu de mots les caracteres, peut-etre artificiels, par 

 lesquels on les distingue'. He noted also that A. purpurascens differed from A. radiatus only in the 

 red stripe of the ligule and that A. valentinus was a short-liguled form of the same. Indeed, A. 

 radiatus and A. purpurascens are synonymous, as also are A. clavatus and A. tomentosus. As is 

 demonstrated below (p. 131), European plants named A. valentinus L. are considered to be 

 possible hybrids between A. homogamos and A. radiatus. 



(3) The activities of the previous period and the start of the new period were punctuated by the 

 contributions of Cassini (1825). When Cassini critically reviewed the existing literature of the 

 Anthemideae, he offered two suggestions to improve the definition of Anacyclus: 'Le vrai genre 

 Anacyclus tel que nous le concevons, differe du vrai genre Anthemis par deux caracteres princi- 

 paux; 1 les ovaires obcomprimes et munis d'une large bordure sur les deux aretes laterales; 

 2 les corolles du disque portant une longue come calleuse, tres remarquable, sur leur divisions 

 interieures'. The main consequences when using these characters were to (i) transfer the Maghreb 

 perennial Anacyclus pyrethrum from Anthemis (as foreshadowed by Link three years earlier 

 (Enum. Hort. Berol. 2 : 344 (1822)), (ii) redescribe the genus, (iii) designate Anthemis valentina L. 

 as the type of the genus, (iv) suggest a division of the genus into two sections: 'dont la premiere 

 seroit caracterisee par les ovaires de la couronne aigrettes et articules avec la corolle, la seconde 

 par les ovaires de la couronne inaigrettes et continus avec la corolle', (v) re-assign the misplaced 

 Anacyclus cretica to a new genus, called Lyonettia Cass., as L. pusilla Cass. It is not clear whether 

 Cassini considered Anacyclus valentinus L. and Anthemis valentina L. to be conspecific, but 

 Loiseleur Deslongchamps (Fl. Gallica : 582 (1807)) clarified the issue when he transferred the latter 

 species to Anacyclus and correctly gave it a new name, Anacyclus radiatus Lois. 



Two major contributions providing synthetic classifications for genera of the Compositae, and 

 the Anthemideae in particular, are found in the works of Lessing and De Candolle. This particular 

 period was characterized by an attempt to give greater consistency to taxonomic groups and thus 

 is marked by a wealth of new names and unusual delimitations. Lessing (1831, 1832) presented 

 one of the more extreme views, for example, when he considered the members of the Anthemideae 

 to belong to a much larger tribe, the Senecionideae. Species of Anacyclus were dispersed into two 

 subtribes, VI Chrysanthemeae and VII Artemisieae. He divided the Chrysanthemeae still further 

 into two groups, the Chrysanthemineae and the Anthemideae, on the basis of the absence or 

 presence of scales (see Humphries, 1976tf). Within group 1 Anthemideae, with receptacular scales, 

 species of Anacyclus were dispersed into two genera. The perennial species, following Linnaeus' 

 classification, was maintained as Anthemis pyrethrum, alongside A. cota L. and A. tinctoria L.; 

 Anacyclus was taken to comprise two species only, A. qfficinarum Hayne and A. radiatus Lois. 



