20 Contributions from the Gray Herbarium 
It seems remarkable that this splendid thistle should have 
escaped notice so long since it grows at the very door, so to speak, 
of one of the principal herbaria of the Pacific coast. An explana- 
tion may be had, however, from the fact that the identity of two 
species in the C. edule group (to which C. praeteriens belongs) has 
been obscure. C. Andrewsii (Gray) Jepson was described from a 
meager specimen showing only an upper branch. When Professor 
Jepson included it in his Flora of W. Middle Cal.ed. 1, 506, (1901) he 
referred toita specimen collected by Miss Eastwood from Tennessee 
Bay with the note “(apparently also at Lake Merced) and distrib- 
uted by her as Carduus amplifolius Greene.” In the second edition, 
p.423 (1911), Tennessee Bay is still given as the only known locality 
and no mention is made of Greene’s amplifolius. I have compared 
Miss Eastwood’s specimens with the type of C. Andrewsii and there 
is no doubt that they are referable to that species; and since they 
agree precisely with the description of Carduus amplifolius Greene, 
Bot. San F. Bay Reg. 217 (1894) I have no hesitancy in regarding 
this name as a synonym. It may be noted that Miss Eastwood's 
material is fine, consisting of several sheets which show the entire 
plant. A distinctive character of C. Andrewsii is the great reduc- 
tion of the upper leaves. These are broadly ovate in outline and, 
as described by Dr. Gray, laciniate-pinnatifid. The lower cauline 
leaves are well-described by Greene, l.c. C. praeteriens is nearest 
this species but is at once distinguished by reason of its compara- 
tively narrow inordinately spinose leaves which are scarcely reduced 
upward and retain to the heads their oblong outline. The narrow 
rigid lobes of the leaves are very different from the broad soft over- 
lapping lobes that characterize C. Andrewsii. The flowers of the 
latter are not‘ whitish ” but roseate (as is still evident even in the 
type) and the heads are smaller (3-4 em.) than those of C. prae- 
teriens. Only two other species of this group, characterized by 
leafy-bracted heads, are known from the vicinity of San Francis¢o, 
viz. C. edule and C. crassicaule and these are obviously very differ- 
ent from our plant. C. crassicaule was first collected by Dr- 
Kellogg in April, 1868 or 1869 and was cited by Dr. Gray, Proe- 
Am. Acad. x. 41 (1875) as representing his new species Cnicus 
quercetorum. He based this species, however, upon an Oakland 
specimen collected by Bolander and this is to be taken as the type 
since his description was evidently drawn from that material rather 





