(148 ) 
In this paper the general divisions follow Bentham and Hooker, 
but there is some deviation in the delimitation of certain genera 
formerly included under Vernonia and Elephantopus. The segre- 
gation followed here is by no means new, but has been practiced 
by various botanists since Lessing, and is abundantly justified by its 
results. The segregated genera, Lachnorhiza, Eremosis, and 
Letboldia in the one case, and Orthopappus and Pseudelephan- 
topus in the other, represent each a natural coherent group of 
species, sharply delimited by the structure of the pappus, the in- 
florescence, or some other morphological characters, and also with 
definite, characteristic geographical distribution. 
he two subtribes, Euvernonieae and Lychnophoreae, while en- 
tirely distinct, nevertheless approach each other in certain genera. 
The genus Vazdzllosmopsis, of the former subtribe, is in particular 
closely similar in inflorescence to Hremanthus of the Lychnophoreae. 
Efremosis and Oliganthes also have heads with few flowers 
more or less closely aggregated in panicles and apparently also 
approaching the Lychnophoreae. It seems probable, therefore, 
that these genera represent most nearly the primitive type of the 
Vernonieae, and that from similar forms the evolution of the tribe 
has proceeded. The structure of the existing genera indicates that 
this evolution has been along at least five different general lines, 
three of them preserving the general feature of aggregated heads 
with few flowers, constituting the Lychnophoreae, and at least two 
tending toward an open inflorescence and many-flowered heads, 
the Euvernonieae. In the latter group one line has resulted in 
the Sparganophoreae, in which the achene is surmounted by a car- 
tilaginous ring. In Sfarganophorus the ring alone is present; in 
Pacourina a row of bristles is within the ring, showing the direc- 
tion from which the genera have come. A second line of evolution 
is shown in the genus Verzonza and its relatives, where the 
changes have been principally in the inflorescence, ranging from 
scorpioid forms with relatively few-flowered heads to paniculate 
forms with frequently many-flowered heads. It is not clear 
whether the subtribes Ethulieae and Stilpnopappeae represent dif- 
erent lines of descent, or whether the differentiation of a pappus 
into forms composed entirely of scales, of similar bristles, or of 
dissimilar series is to be regarded as indicating distinct lines of 
* Engler and Prantl, Die Natirlichen PAanzenfamilien 45: 120-131. 1894. 
