53 
OBSERVATIONS ON THE COMPOSITA.—IIL 
By Epwarp L. GREENE. 
EUPATORIACEE (continued). 
Tue old and typical species of EvpaTorium being tolerably 
well at agreement in habit, and exhibiting uniformly a 
5-angled achene, the angles not being obscured by inter- . 
vening secondary ribs or striw, the recognition, early in the 
present century, of one Mexican, and one Atlantic North 
American species of this alliance having achenes 10-striate, 
and wearing an aspect somewhat unlike that of Hupatorium, 
suggested to several botanists independently, the idea of 
establishing a new genus on this foundation; and so were 
added to the nomenclature of the group CoLEOSANTHUs (Cass. 
Bull. Philom. 67. 1816), and its synonyms, Rosalesia (Llav. 
& Lex. 1824), Brickellia (Ell. 1824), Wikstreemia (Spreng, 
1826), and Bulbostylis (DC. 1836). Here we have five 
eminent botanists engaged independently in the consideration 
of a given type, and each one reaching the same conclusion, 
namely, that the plants represent a genus distinct from 
Eupatorium; while Sir William Hooker, into whose hands 
had fallen another member of the group from the north- 
western part of North America, ventured to give it a specific 
name under Eupatorium (E.? grandiflorum, Hook. Fl. ii. 26) 
only with a mark of doubt as to its being truly of that genus. 
The old axiom in systematic botany, that the genus deter- 
mines its character, and not the character the genus, implies 
that the recognition of a genus is the first thing, and that its 
characters, few or many, feeble or strong, are an after con- 
sideration. So that under this axiom, a genus may hardly 
have a stronger warrant than the fact that six or seven com- 
petent botanists have independently given it recognition. 
It were therefore presumptuous in a mediocre botanist to deny 
the validity of Coleosanthus, even if its only character be the 
10-striate achene. But there are really some other differ- 
EryrsEa. Vol. I, No. 3 [1 March, 1893]. 
