iqi61 SHERFF—BIDENS 501 



foliage. The immature type of B. erithmifolia is matched exactly 

 by J . Triana 1374 (ibid.), a specimen from Bogota, Colombia, which 

 is superior in showing not only flowering heads but also numerous 

 achenes. These achenes are mainly 2-aristate, but some are 

 3-aristate and so agree perfectly with achenes of B. delphinifolia. 6 

 Fortunately, I have found another specimen by Triana of the same 

 number in the British Museum, and this shows the slightly different 

 foliage of B. delphinifolia, thus removing all doubt as to the identity 

 of B. delphinifolia with B. erithmifolia. 



This species is exceedingly variable in foliage. Triana 1375 (in 

 Herb. Brit. Mus.) shows one plant with leaves tripartite, the leaf- 

 lets being incisely dentate, and another plant with simple, ovate- 

 lanceolate leaves. In fact, the sheet of Triana 1374 in Paris bears 

 another specimen (beside that cited above) which has similarly 

 simple, ovate-lanceolate leaves, showing that in the field Triana 

 considered the simple leaves and the finely divided leaves as belong- 

 ing to the same species. More recently, further material has come 

 from Colombia {Herbert II. Smith 1980, San Lorenzo ridges, Santa 

 Marta, in Herb. Field Mus., Herb. New York Bot. Card., etc.) 

 which shows many leaves like those of the type of B. erithmifolia, 

 but with leaf divisions narrower. A range of leaf outlines thus is 

 shown that seems unbelievable, easily entitling B. erithmifolia to 

 rank in this respect with such species as B. heterophylla Ort. 



Bidens hirtella H.B.K. Nov. Gen. 4:232. 1820. — Bidens pro- 

 cumbens H.B.K. loc. cit. 



The type of B. procumbens (in Herb. Mus. Hist. Nat. Paris) 

 differs only in the slightest way from that of B. hirtella {ibid.). 

 Kunth (H.B.K. loc. cit.) admitted the two forms to be very close. 

 Had he possessed the wide range of data concerning variations in 

 the related species of Bidens that, during the century since then, 

 have steadily accumulated, he would not have hesitated to treat 

 these as specifically the same. 



Bidens chrysanthemifolia (H.B.K.). comb. nov. — Cosmos chry- 

 santhemifolia H.B.K. Nov. Gen. 4:239. 1820; Cosmea chrysan- 

 themifolia Sprengel, Syst. Veg. ed. 16 1 ". 615. 1826; Cosmos 



6 It is interesting to note that the mature head on the type of B. delphinifolia has 

 at least one achene that is positively 2-aristate, showing no indication of a third awn 

 ever having been present (cf. H.B.K. loc. cit., "triaristata"). 



