44 ERYTHEA. 
but this is a tribal, or subordinal characteristic, and not a 
special mark of the genus Hupatorium. 
Modern authors, beginning with Cassini or with Lessing, 
record a 4 or 5-angled achene having smooth sides, and being 
surmounted by an uniserial pappus of scabrous bristles, as 
the essential characters of this type genus; to which, how- 
ever are added, as of less importance, an imbricated involucre 
of a considerable number of unequal bracts, and a corolla the 
limb of which is not companulate, but funnelform, 7. e. atten- 
uate to the tube. I have long been thinking that, in this vast 
family of the Composite, where the flowers are always con- 
gested into dense heads, any diversity in the form of the 
corolla should be more than usually significant to the 
systematist. The natural tendency of this crowding is to 
reduce all shapes of the corolla to one; and the difference 
between a funnelform limb and a campanulate limb, when all 
are alike crowded, may signify much genetically. WILLUGBEYA 
(Neck. Elem. 82. 1790—Mikania Willd. Sp. iii. 1481. 1803) 
was taken out of the Eupatorium of Linneus, on account of 
its 4-bracted involucre, 4-flowered heads, and campanulate- 
dilated corolla. Baillon restores the species to Hupatorium, 
induced no doubt by the consideration that some had been 
admitted into Willugbeeya having the funnelform corollas of 
Eupatorium. These, it is true, had been placed in the rank 
of a genus Kanimia (Gardn.; Hook. Lond. Journ. vi. 446. 
1849); and it is now claimed that Willugbwya and Kanimia 
differ both from Hupatorium and from each other in the form 
of the style-branches. These are to be claviform (therefore 
terete and obtuse) in Hupatorium, acutish in Willugbeya, 
and obtuse but compressed in Kanimia. That the few and 
definite involucral bracts and flowers has not been thought a 
character of the highest import is evident from the fact that 
species with 3-flowered and even 1-flowered involucres are 
retained in Hwpatorium, by those who exclude from it the 
genera last mentioned. Hence the greatest stress appears 
to be laid upon characters of the corolla and of the style- 
branches. That nearly all the species of the two lesser 
