10.5 
natural Family of Flants called Compositce. 
But if these four plants, so extremely different from each other 
in pappus and form of the pericarpium, really belong to the same 
genus, as their habit seems strongly to indicate, there can be no 
reason to separate from them Alcina of Cavanilles, erroneously 
considered by Willdenow as a species of IVedelia : and Dysodium 
of Richard, published in Persoon’s Synopsis, though differing 
from all the others in the form of its pericarpium and in that 
of its receptacle, must also be reduced to this genus. If, how¬ 
ever, the part described by Linneus as pappus in Melampodium 
americanum be really such, and if the pericarpium itself vary so 
widely both in form and surface, it would be inconsistent with 
% 
the principles of division generally adopted in Composite, to 
unite all these plants into one genus, notwithstanding their great 
resemblance in habit as well as in the other parts of fructifica¬ 
tion ; and it would be at least in vain to look for any combining 
character in this part of their structure. 
A careful examination of the female flowers, especially in an 
early stage, removes this difficulty, by proving that the supposed 
external coat of the ovarium, with its various inequalities of sur¬ 
face, some of which have been described as pappus, is in reality 
an involute bractea or foliolum of the involucrum, like that of 
Micropus , completely inclosing the ovarium, but from which in 
several species of the genus it is entirely, and in others in great 
part, distinct. 
Craspedia 
first appears in Forster’s Prodromus Florulae Insularum Austra- 
lium, where an essential generic character is given, but no de¬ 
scription of the species. The genus is adopted and the cha¬ 
racter received without remark by Willdenow in his edition of 
Species Plantarum, and by Persoon in his Synopsis. Among 
George Forster’s drawings of subjects of natural history made in 
vol. xii. p Cook’s 
p 
