FLORA OF BRAZIL. 429 
3805. Elephantopus Martii, Graham. 
3263. Trichospira biaristata, Less. 
EuraroriAcezx, Less. 
If the characters of the genera which constitute the divi- 
sion dlomiee of the subtribe Eupatoriee of De Candolle be 
compared with each other, they will be found, with the 
single exception of Isocarpha, which is a good genus, to 
differ in no essential respect. Such differences as a single 
ordouble row of involucral scales, few or many florets in 
the capitula, and pilose or glabrous corolle, surely cannot 
be considered as sufficient generic distinctions. Were such 
principles to be acted on throughout the order, we should 
find that Eupatorium alone would resolve itself into at 
least half a dozen genera. I therefore propose that the 
Senera alluded to should all be thrown into a single one 
under the original name of Piqueria, the naked recep- 
tacle and the absence of pappus constituting its essen- 
tial character. It may be divided into two sections, the 
first, Eupiqueria, containing the few flowered, and the second, 
Alomia, the many flowered species. : 
It must have occurred to all those who have made the 
“omposite their study, that this is not the only instance in 
which De Candolle has multiplied genera to a most unwar- 
Tantable extent, Though very much is due to the learned 
Genevese ` Botanist for his great labours on the order, more 
"specially in bringing together the great mass of species, 
Which Previous to his investigation of them were scattered 
Man undigested form through many works, yet the laxity 
[MF his views with ‘regard to genera, the want of a better 
Principle than that of the nature of the style on which to — 
‘ound the great divisions, and the vast additions which have 
1? made to it during the last ten years from all parts of 
the World, render a revision of the whole tribe much called - 
for by some’ one who has the time, the talents’ and the 
Materials for such a task. sas: si beyisask ^ 
