Blake — A Revision of the Genus Viguiera 15 
counterpart in this respect among the known species of Viguiera, 
and but slightly approached by V. laciniata Gray. The alliance 
of these species is indicated primarily by their involucres, of 3—4- 
seriate graduated lance-subulate phyllaries with slightly broadened 
indurated ribbed and vittate base and long loose linear herbaceous 
apex, which are thus identical in all essentials with those of the 
series Dentatae of Mexico and Lower California, to which V. lacin- 
tata above mentioned belongs. The three species of Gymnolomia 
just mentioned are found from Texas to Oaxaca, where the only 
species of the Dentatae occurring is V. dentata itself, an herb with 
entire or merely toothed leaves. It seems probable, therefore, that 
these species of Gymnolomia have been derived from some group of 
Viguiera now extinct. The next five species of Gymnolomia in 
Robinson & Greenman’s enumeration, including the curiously iso- 
lated G. Porteri of Stone Mountain and vicinity, Georgia, with the 
later described G. brevifolia Greene, G@. nevadensis A. Nelson, and 
G. obscura Blake, form a section which may be called, from its 
oldest known species, the G. multiflora group. They have no pre- 
cise analogues in Viguiera, but their relationships are made clear 
by their receptacular pales, which have the peculiarly stiff strongly 
ribbed base and rather abrupt firm mucro of the series Dentatae of 
Viguiera. Their involucres also are of the same general type found 
in that series, except that the indurated and ribbed base of the 
phyllaries is practically wanting. It is probable that the G. multi- 
flora group, like the G. tenuifolia group, has been derived from a 
group of Viguiera most nearly related to the Dentatae but now 
extinct. 
The relationship of Gymnolomia microcephala Less., G. guatema- 
lensis (Rob. & Greenm.) Greenm., and G. costaricensis Benth. to 
species of the genus Hymenostephium has already been discussed, 
and it must suffice to repeat here that not the slightest difference 
can be found between G. costaricensis and H. cordatum aside from 
the presence or absence of pappus and the concomitant presence 
or absence of hairs on the achene, while the differences in other 
characters between the other pairs of species are only of varietal 
value. G.? subfleruosa (H. & A.) B. & H. of Robinson and Green- 
* Blake, Journ. Bot. liii. 268 (1915). 
