GNAPHALIUM AND XERANTHEMUM. 53T 
lies, that they might readily be mistaken for the stigmasy 
which are far less conspicuous. This is also the case with 
the Carlina echinus, the Echinus mutans of Cassini ; but 
in it the rays of the pappus are unconnected at the base> 
and the figure of the florets, as in Cnkus, Carduus, &c. 
In the Carlina vulgaris and corymbosa, the rays of the 
pappus end in a naked thorny point. The InulecB M. 
Cassini, limited to Inula and Coryza, will possess sufficient 
marks in their herbaceous involucrum, and in haying ligu- 
late florets in the ray, to keep them distinct. Bupthalmum, 
referred by M. Cassini to Inulew, belongs to the Helianr 
thecB, and Carpesium to the Anthemoidea, 
We have seen, therefore, the intimate affinity that exists 
between the true Carduacees and the plants referred to 
Gnaphalium and Xeranthemum of Linnseus. In compar- 
ing the EupatorecB of Cassini, to which the VernoniacecB rof 
the same distinguished botanist must be reduced, with the 
Carduacece, we shall find the affinity almost equally inti- 
mate. The former group differs from the latter in the ab- 
sence of setae at the base of the anthers. In Liatris and 
Pteronia^ which belong to the EupaiorecB, the involucrum 
is analogous, especially in Liatris elegans, vv^here it is fur- 
nished with long, lanceolate, acute, purple rays, like those 
of Xeranthemum. In Liatris also, the rays of the feathered 
seed-crown are united at the base ; and both it and Pteronia 
have lanceolate stigmas analogous to those of some species 
of Serratula. If we turn our investigation to the Labiati- 
Jlorce, we shall meet with similar results. Let us compare, 
for example, Xeraiithemum, in which the marginal florets 
are evidently bilabiate, the anthers bicaudate at the base, 
and we shall find how closely these characters bring it to 
the LahiatiflorcE The genus Catananche, which belongs 
• In some of the normal group of Carduacece, we find the limb of the 
florets irregular, — and in Carduus Marianus^ the genus Silybum of Maench, 
it is evidently bilabiate. 
