26 ARKIV FOR BOTANIK. BAND 13. N:0 15. 



Vernonia gracilis H. B. K. is a plant common in the 

 dry plains of Columbia. Like all Vernonise it is highly vari- 

 able, especially as to the pubescence of the leaves and the length 

 of the subulate tips of the scales. V. Moritziana Sch.-Bip. 

 was described upon a form with very short tips. The genuine 

 V. gracilis has leaves nearly glabrous beneath. I have based 

 the new subspecies upon specimens having the leaves densely 

 floccose-tomentose beneath. Just because of this pubes- 

 cence ^Hieronymus took the same plant to be V. Cotoneaster 

 (W.) Less., a species from Bahia, differing from V. gracilis 

 in its smaller involucres. The material at hand of the sub- 

 species is too meagre to give a reliable idea of its systematic 

 rank; perhaps it is a distinct species. 



It is puzzling that V. gracilis subsp. tomentosa has been 

 found to occur in Bequia, far away from its proper region, 

 which is Columbia. 



Vernonia tricephala Gardn. 



Tabula nostra VI, fig. 5 (cyma). 



Vernonia gracilis H. B. K. var. villosa Lessing, 1829, p. 303; De 

 Candolle, 1836, p. 50. 



Vernonia tricephala Gardner, 1846, p. 223; Baker, 1873, p. 68. 

 Vernonia tricholepis Grisebach, 1861, p. 354, p. p. non DC. 

 Cacalia tricephala Kuntze, 1891, p. 971. 



Pappi setae interiorcs persistentes, numero 30, tenues, 

 filiformes, albidse, quam exteriores distinctse circ. 4-plo lon- 

 giores. Corolla 5 mm longa, tubo sensim dilatato long. 2 / 3 

 corollse, glandulis nonnullis obsito, limbi laciniis dorso pilis 

 longis et glandulis globuliformibus instructis. Antherse 2 

 mm longae, ligula i / 5 long, antherse, auriculis acutis. 



Hab. in Trinidad: Crueger, sine loco ace. (G); Cedros, 

 loco sicco, 15. 1. 08, Broadway n. 2174 (DC, KU). 



Specimens from Trinidad agree exactly with the genuine 

 V. tricephala from Brazil. 



Grisebach cites, 1861, p. 354, two West Indian plants 

 as belonging to V. tricholepis DC. Neither of them belongs 

 to that species, the former, Jamaica, Purdie, being V. se- 

 ricea L. C. Rich., certainty not collected in Jamaica, the second 

 is just V. tricephala Gardn. 



