TROPICAL AMERICAN COMPOSITAE. 15 



acute mucronatis plerisque ca. 5 mm. longis; corollis viridescenti- 

 flavidulis tubulosis gradatim paullulo sursum ampliatis 3.5 mm. 

 longis; dentibus limbi 5 breviter deltoideis extiis granulatis; antheris 

 apice breviter crassiusculeque appendiculatis; styli ramis filiformibus 

 paullo ad apicem incrassatis et nigrescentibus, achaeniis griseis 2.7 mm. 

 longis deorsum decrescentibus angulis hispidulis; pappi setis ca. 32 

 corollam subaequantibus albis vix barbellatis. — Peru: woods near 

 a river, Caraz, Dept. Ancachs, alt. 2200 m., May 21, 1903, ]]'cbcr- 

 baiier, no. 3027 (Berl., phot, and fragm. Gr.). 



In many respects similar to E. solidaginoidcs HBK. but not as yet 

 connected by intermediates and rather too different in appearance to 

 be regarded as a variety until intergradation has been demonstrated. 

 In E. solidaginoides the branches of the inflorescence present a 

 somewhat racemose appearance, the heads being rather evenly dis- 

 tributed along them. In E. flexile this is not the case, the branches 

 being floriferous chiefly toward the tip. In E. solidaginoides the 

 heads are 10-15-flowered and the involucre (when fresh or softened 

 by boiling) is rather narrowly campanulate, the scales being thin. 

 In E. flexile the heads are about 20-flowered, the involucre broadly 

 campanulate, and the scales somewhat firmer in texture. The 

 achenes of E. solidaginoides are from 1.8-2.1 mm. long, while in E. 

 flexile they are about 2.7 mm. long. 



E. FULiGiNOSUM HBK. Nov. Gen et Spec. iv. 110 (1820); Robinson, 

 Proc. Am. Acad. liv. 302 (1918). Of this little known Colombian 

 species further material is now at hand extending its recorded dis- 

 tribution to the department of Huila, where these specimens were 

 collected in forests on the Cordillera Oriental, east of Neiva, Aug. 

 1-8, 1917, by Drs. Rusby & Pennell, nos. 573 (N.Y.) and 977 (N.Y.). 

 In all characters these correspond closely with a photograph (Gr.) of 

 the type (Par.) and of their specific identity there can be no doubt. 

 However, the receptacle bears a short, thin and very fugacious pubes- 

 cence which commonly disappears so completely with the fall of the 

 achenes as to leave the receptacle entirely glabrous and merely 

 punctate, as it was described in the original diagnosis of the species. 

 Technically, having a hairy receptacle, E. fuliginosum must be trans- 

 ferred to § Hebeclinium. In practice, however, it will be well to retain 

 at least a cross-reference to it in § Subimbricatum, for the differential 

 character from its obscurity and fugacious nature is here very likely 

 to be overlooked. The labels of the material at hand record the 

 plant as a tall shrub with greenish-yellow flowers. 



E. (§ Subimbricata) Gascae, spec, nov., dense villoso-tomento- 



