Macbride — Reclassified or new Spermatophytes 37° 
exhibits it to greater or less degree. The variety brevibracteatus 
was originally based on a plant with fewer flowers in the heads 
and greatly abbreviated bracts and in typical form is known only 
from Kern County. Specimens from this region show great vari- 
ation, however, and confirm the opinion I expressed in 1917 
(Contrib. Gray Herb. xlix. 59) that the variety brevibracteatus is 
intermediate between the typical form and the variety filifolius, 
and is evidence, therefore, that but a single specific unit is here 
concerned. Z 
Cordylanthus littoralis (Ferris), comb. nov. Adenostegia littor- 
alis Ferris, Bull. Torr. Club, xlv. 413 (1918). 
CoRDYLANTHUS RAMOsUS Nutt., var. puberulus, var. nov., 
ubique solum puberulus, bracteis haud ciliatis. — Northern Colo- 
rado to Wyoming, Oregon and Nevada. — Nevapa: sagebrush 
lands, Mountain City, Elko Co., Aug. 13, 1912, Nelson & Mac- 
bride, 2197 (rypE, Gray Herb.). 
Rydberg based his Adenostegia ciliosa Rydb. Bull. Torr. Club, 
xxxiv. 35 (1907), on a Tweedy specimen from western Wyoming 
which I have not seen but I feel confident I am correct in referring 
it to typical C. ramosus which was described originally, DC. Prod. 
x. 597 (1846), as having “ ciliate bracts” and furthermore the 
bracts of a scrap in the Gray Herbarium, marked ‘“ C. ramosus ”’ 
and purporting to be part of the type are, indeed, distinctly ciliate. 
The bracts of the variety are merely puberulent and it is the ex- 
treme development of ciliation in the typical form of the species 
. 88 represented by the type, by Leiberg’s number 848 and by Mac- 
bride & Payson’s number 3840 that induces me to give the more 
Common and merely puberulent form even varietal recognition 
for there appears to be no other character, such as a longer calyx 
4s Rydberg’s description implies, that is correlated with the ab- 
Sence of cilia. 
Mrs. Ferris accepted Rydberg’s assertion, Bull. Torr. Club, xl. 
484 (1913), that C. bicolor A. Nels. “is evidently the same as 
Adenostegia ciliosa Rydb.” and since C. bicolor is exactly C. capi- 
tatus (which has two stamens and 1-celled anthers) she has very 
naturally referred C. ciliosus to the same species, notwithstanding 
the fact that it has, of course, the four stamens and 2-celled 
anthers that characterize C. ramosus! Rydberg’s superficial re- 
duction of @. bicolor A. Nels. has thus caused considerable misappli- 
+ 
