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



for instance, in St. Thomas. This supposition is made more 

 probable by the fact that another Vernonia, V. albicaulis, 

 which certainly does not live in Hispaniola, is in the 

 Berlin Herbarium labelled: Mayerhoff, Sto. Domingo in 

 1852. It represents the form of V. albicaulis common in 

 St. Thomas. 



The type specimen of V. racemosa Delponte has not been 

 examined. Since the description of this species agrees exactly 

 with the Hispaniolan plant, and the type was collected there 

 (by Bertero), there can be no doubt as to its identity. 



In the Lamarck Herbarium at Paris there is a specimen 

 of the subspecies labelled: Conyza de Mr. De Badier, de la 

 Guadeloupe, n. 44 . In the Prodromus Herbarium, however, 

 the same label is affixed to V. arborescens Sw. (3. ovatifolia 

 DC, i. e. to V. albicaulis Pers. a fact illustrating how plants 

 have been confused in ancient times. 



5. The allies of Vermonia segregata Gleason. 

 Vernonia segregata Gleason. 



Tabula nostra V, fig. 3 (inflorescentia). 

 Vernonia segregata Gleason, 1913, p. 327. 



Hab. in Cuba: Oriente, Camp la Gloria, prope Sierra 

 Moa, 2430. 12. 10, Shafer n. 8051 (KU). 



I have seen only very imperfect portions of this puzzling 

 species. It is readily recognized by its peculiar inflorescense : 

 capitate clusters of heads subtended by bracteal leaves, the 

 clusters both terminal and axillar. As to the leaves and the 

 involucres it recalls the species near to V. havanensis DC, viz. 

 V. hieracioides and V. cubensis, but differs widely in the in- 

 florescence. 



Subsect. Scorpioides. 



Frutices elati, sazpe scandentes. Folia lata, scepe magna, 

 basi scepe contracto-attenata. Inflorescentia cymoso-scorpioidea, 

 cymis elongatis. midtifloris, simplicibus vel parce dichotome 

 ramosis, calathidiis sessililnis. contiguis; hravteis nullis. Ca- 



