286 ROBINSON. 



mostly obtuse or rounded tip, serrate to undulate or subentire, 4-6 cm. 

 long, 2-3 cm. wide, 3-nerved from above the base; this cuneately 

 decurrent on the petiole; heads about 5-flo\vered, slender, shortly 

 pedicelled in open forking c;^Tnose panicles; involucral scales 11-16, 

 stramineous and subscarious, very unequal, rather loosely imbricated, 

 acute; corollas pale yellowish (Smith); achenes pubescent. — Nov. 

 Gen. et. Spec. iv. 106, t. 340 (1820) ; Klatt in Engl. Bot. Jahrb. viii. 

 34 (1887); Heering, Mem. Soc. neuchat, Sci. Nat. v. 419 (1913). 

 E. iresinoides, var. a. villosum Steetz in Seem. Bot. Herald, 145 (1854); 

 Hieron. in Engl. Bot. Jahrb. xxviii. 573 (1901). E. celosioides Willd. 

 ex Steetz, 1. c. (invalid name, needlessly published). E. celtidifolium 

 Klatt, 1. c, not Lam. 



Magdalena: Las Nubes and Onaca, alt. about 600 m., H. H. Smith, no. 

 503 (Gr., U. S.); lower hills between Rio Frio and S. Andres de la Sierra, alt. 

 about 200 m., Pittier, no. 1716 (U. S.). 



Touma: at the base of the Andes near Ibague, Humboldt & Bonyland. 



Huila: east of Neiva, Rushy & Pennell, no. 1048 (X. Y.); Patico, Lehmann, 

 no. 4768 (N. Y.). 



El Valle: among shrubs on savannahs about Tocotd, alt. 1600 m., Leh- 

 mann, no. 3430 (Gr.); on stony sterUe soil along the Rio Dagua, Lehmann, 

 no. 3813 (Gr.); La Paila, Holton, no. 320 (Gr., K.). 



[Venez., Margarita, St. Vincent, Martinique, Panama. Said by Hemsl. 

 Biol. Cent.-Am. Bot. ii. 96 (1881) to extend southwanl to Peru.] 



Var. a. viUosum Steetz, 1. c. Copiously pubescent; leaves softly 

 and conspicuously tomentose beneath. — Synon. and distrib. as above. 



Var. (3. glabrescens Steetz. 1. c. Finely and often rather incon- 

 spicuously pubescent. 



Magdalena: on rocky hills by the seashore near Plaza Brava, H. H. Smith, 

 no. 607 (Gr.). 



[Panama, Venez., Trinidad.] 



Poorly marked and of merely formal A^alue. 



The flowers of this species are sometimes reported on field-labels 

 as yellow. Since, howcAcr, really yellow flowers are decidedly rare 

 in the Evpatoricac it is not unlikely that they are here nearly or quite 

 white and that the impression of yellow color arises from the stramine- 

 ous scales, which in this species are considerably more conspicuous 

 than the minute and almost included florets. 



26. E. salicinum Lam. Shnib; branches .somewhat angular, 

 puberulent when yoinig; leaves opposite, subsessile, oblong-lanceolate, 



