WESTERN TROPICAL AMERICA. 117 
453. Brpens leucantha, Willd.—DC. Prod. v. 5. p. 598.—Tepic. 
454. Brens hispida, Humb. et Kunth.—DC. Prod. v. 5. p. 599.—Peyta, in Columbia, 
455. Cosmos caudatus, Humb. et Kunth.—DC. Prod. v. 5. p. 606.—Bidens bipinnata, Hook. 
et Arn. Bot. Beech. p. 436. non Linn.—Various stations, from Tepie to Guayaquil. 
456. Cosmos carvifolius, sp. n., glaber, foliis bipinnatipartitis lobis capillaceo-linearibus inte- 
gerrimis, involucri squamis exterioribus circa 8 lanceolatis acutis quam interiores brevioribus, 
acheeniis glabris levibus longiuscule rostratis 5-aristatis—Tepic. 
In omnibus cum descriptione Kunthii C. parviflori convenit, nisi acheeniis longius rostratis ad angulos non 
scabridis et aristis in omnibus quos vidi capitulis semper quinque nec tres, 
457. Lirocu#{TE macrocephala, Hook. et Arn. Bot. Beech. p. 436. (sp. n.)—Acapulco. 
458. Lirocuz£tTE umbellata, DC. Prod. v. 5. p. 610.—Tepic. 
459. Dunantia achyranthes, DC. Prod. v. 5. p. 627.—Deless, Ic. Sel. v. 4. t. 37. — Gulf of 
Fonseca. 
The style of this plant, very well represented in the figure quoted, as well as 
that of Isocarpha oppositifolia, Br. (which I have above shown to be a congener), 
and of some other Senecionide, is, strictly speaking, nearer that of Vernoniacee, 
and shows the difficulty of relying on the style alone for the distinction of these 
tribes. In both tribes the sterile summit of the branches is hispid, and neither 
club-shaped as in Eupatoriacee, nor (except in a few Senecionide) broad or 
flattened as in Astere@. In Vernoniacee the hairs usually extend on the outside, 
much below the end of the stigmatic leaves, and even below the fork of the style, 
without any sudden difference in the length of the hairs, and the sterile part is 
always long and pointed ; in the greater number of Senecionide, on the contrary, the 
branches of the style are usually entirely smooth up to the end of the stigmatic 
lines, and are there either truncate, with the sterile part reduced to a tuft of hairs, 
or terminated by a conical or elongated and pointed sterile part, with the hairs at 
its base longer than the others ; but there are many Vernoniacee where the stigmatic 
portion of the branches is nearly or even perfectly smooth, and there are several © 
Senecionide, like Dunantia, in which the long hairs at the base of the sterile 
portion are wanting, and the stigmatic part is not wholly smooth on the outside. 
In such cases other characters must be resorted to for the distinction of these two_ 
large groups, which are, generally speaking, natural, and would be still more so if 
the opposite-leaved radiate, Vernoniacee (Pectidee and Liabee), could be trans- 
ferred to Senecionide. 
460. Satmea angustifolia, sp. n., foliis oblongo-lanceolatis utrinque angustatis acutis cauleque 
scaberrimis, corollis rectiusculis, aristis eequalibus.— Mexican coast. 
Frutex, ramulis teretibus pube brevi albida seabris. Folia subsessilia, 3. poll. longa, medio 6-9 lin. lata, 
HH 
