140 THE JOURNAL OF BOTANY 
0-6 cm. long. Corolle in toto 0-6 em., harum lobi 0-3 em. long. 
Antherw 0°35 cm. long. s 0-4 cm., rami 0-25 cm. long. 
Acheenia cruda modo 0 ‘6 em. long. 
The characteristic points of the species above proposed contrast 
as follows :— 
Branches compact. Leaves obovate or oblanceolate, at 
' least 5 mm. long. Heads 4-5-flosculed. Tube of 
corolla nearly six times as long as the lobes. Anthers 
tyle-arms glabrous . ; : » €, conocephala, 
Branches compact. Leaves oblong, not exceeding } mm. 
long. .Heads 1-2-floseuled. Tube of corolla about 
three times as long as the lobes. Anthers connate. 
- Style-arms glabrous ; ; ‘ : - ©. microphylla. 
Branches divaricate. Leaves linear. Heads 5-flosculed. 
Tube of corolla not longer than the lobes. Anthers 
connate. Style-arms puberulous at the back , C. subspinescens, 
The genus is a very interesting and curious one. Originall 
referred by Mueller to Eurybia 
hesitation that Bentham followed this lead. He writes (loc. eit.), 
un 
e case. Notwithstanding these anomalies, as I know of no 
genus to which it is more nearly allied, I have left it in Olearia ag 
in a genus by itself a plant which, as he s oWs, possesses such 
tion of the Composite in the Genera Plantarum (vol. ii. p. 227). 
Plainly Olearia is not the place for these plants, as, leaving out the 
Aomogamous heads, not only are the style-arms emphatically not 
those of Asteroidea, but, contrary to what Bentham says, I find the 
anthers of C. conocephala by crear at the base, though very 
it is true. A i 
8 
The three plants are undoubtedly congeneric, though, as shown 
capitula is so different. 
: r much cogitation, I conclude that Cratystylis should be 
included among the Inuloidea, though without being able to fix 
upon an affinity for it in that tribe. It may be added that but for 
the flattened style-arms, a Vernoniaceous affinity might be defended, 
and perhaps with more than plausibility. In all three the styles 
are thickened (hence the generic name—Gr. xparis, strong), and 
