JOHNSTON. — A REVISION OF THE GENUS FLAVERIA. 281 



As he concluded both that the plant he had at hand was not a Milleria, 

 and that Flaveria was not a good genus, he proposed a new name, 

 Brotera, so that the plant stood as Brctera Contrayerva, Spreng. The 

 plant described and illustrated by Sprengel is, however, an entirely 

 different plant from Milleria Gontrayerba, Cav. Wildenow (1804) in 

 Species Plantarum, iii. pt. 3, 2393, having previously used the name 

 Brotera for a genus (Cardopatium, Juss.) published Nauemburgia triner- 

 vata for Sprengel's plant, not for Cavanilles's. Lagasca, however, Gen. 

 et Sp. Nov. 33, no. 406 (1816), named a plant Flaveria repanda, 

 which Sprengel in 1826, Systema Vegetabilium, iii. 500, identified with 

 both Nauemburgia and Brotera. Sprengel also enumerated F. Gon- 

 trayerba, Pers., F. angustifolia, Pers., and F. linearis, Lag. As late 

 as 1832, Lessing, Synopsis Generum Compositarum, 235, maintained 

 Nauemburgia distinct from Flaveria, and DeCandolle, Prod. v. 635 

 (1836), retained Broteroa (Brotera, Spr.) trinervata. As the distinc- 

 tion between Brotera (or its synonym Nauemburgia) and Flaveria con- 

 sists merely in the presence of setae upon the receptacle, a character 

 variable in some genera of the Compositae, it alone is not sufficient 

 to separate the two. As other characteristics of the plants correspond 

 very well, it has seemed best to unite the two genera in this revision. 



Since the publication of the above species, as has been said, seven good 

 species have been added to the genus, and about as many more plants have 

 been given new names under Flaveria, which have subsequently proved 

 identical with existing species or not to belong to the genus at all. The 

 abundance of material at hand has afforded opportunity for better char- 

 acterizing the species, for increasing the known range of some of them, 

 and it has also furnished sufficient evidence for naming one variety and 

 four new species of plants which have hitherto been placed with others. 



It may be said that the genus groups itself fairly well into subdivisions ; 

 for example, F. australasica and F. repanda are similar in habit, and are 

 the only two having setae upon the receptacle. Those whose heads have 

 three bracts also form a characteristic group, which, however, passes into 

 the group characterized by five bracts. Besides those with perennial 

 roots which do not resemble each other at all, there are several other ex- 

 ceptional forms, as F. anomala, which has the three bracts with bulbous 

 bases, and F. chloraefolia, which is the only species with conspicuously 

 perfoliate leaves. In the subgroup, which is characterized by possessing 

 three involucral bracts, however, there has been considerable confusion in 

 separating the species, due to the similarity sometimes in habit and again 

 in floral structure. F. chilensis is the only one of this group having a 



