328 Gleason: Studies on West Indian Vernonieae 



brown and finely tomentulose ; leaves numerous, crowded, firm, 

 dark green, oblong to elliptic-obovate, broadest at or above the 

 middle, the principal ones 2.5-4 cm - l° n g an d I-I-5 cm. wide, 

 obtuse or subacute, entire, obtuse or rounded at the base, scabrel- 

 late and resinous-punctate above, glabrous beneath and densely 

 punctate with resinous globules and impressed black glands; 

 midvein puberulent, the lateral veins inconspicuous; petioles 1-2 

 mm. long; inflorescence terminal, irregular in shape, consisting 

 of several short, simple or sparingly branched cymes 2-4 cm. long, 

 naked below, and bearing 2-8 crowded heads in a terminal sub- 

 capitate cluster, or at the base of the branches; bracteal leaves 

 1-3 subtending each cluster of heads, resembling the cauline in 

 shape, 5-15 mm. long; heads about 8-flowered; corollas white; 

 involucre campanulate, 3-4 mm. high, its scales loosely and ir- 

 regularly imbricated, appressed at the base, but spreading at the 

 tip, stiff and firm in texture, the outer narrowly triangular-lanceo- 

 late, long-acuminate, the inner narrowly oblong-linear, tapering 

 gradually to the acuminate puberulent apex; achenes thinly 

 pubescent, 2 mm. long; pappus nearly white, the outer series 1 

 mm., the inner 4 mm. long. 



Type, Shafer 4050, from rocky river banks in the vicinity of 

 Camp San Benito, Oriente, Cuba, altitude 900 m., February 24, 

 1910, deposited in the Herbarium of the New York Botanical 

 Garden. Other sheets in the same herbarium, all collected in the 

 mountains of Oriente, are Shafer 8051, stated to be vinelike and 8 

 feet high ; Shafer 8216, 1 .5-2 feet high ; and Shafer 4446, described 

 by the collector as an herb three feet high with purple flowers. 

 Notwithstanding field differences in the color of flowers or texture 

 of stem, all four numbers clearly belong to the same species. 



The relationship of V. segregata is puzzling. The subcapitate 

 clusters clearly represent a modification of a scorpioid type, and 

 most closely resemble the inflorescence of the Scorpioideae aggre- 

 gatae. The involucre is quite different, however, from that of 

 typical members of the group. For the present, it has been con- 

 sidered advisable not to assign the species to any group. 



Species-group Havanenses 



In recent work on Vernonia, V. havanensis and V. Ottonis have 



been considered identical (Revision, 192). The large series of 



specimens now available for study permits the ready separation 



of two species, with characters so typical that to each can be 



