36 Muhlenbergia, Volume 4 
ler linear divisions and lobes. None of the specimens in the 
collections examined come accurately under this description. 
Hall's 2161 from the San Jacinto mountains has the foliage as 
described, but the flowers, though smaller than those of typical 
Parryz, more resemble those of that species. Parish first discoy- 
ered this species at Palm Springs, on the Colorado desert. A 
specimen in his herbarium, No. 6074 from West Canyon, Palm 
Springs, is from the type locality. The flowers conform to 
Gray’s description, but the leaves do not. The lower leaves, 
rounded in outline, have the main (3 or 5) divisions cuneate and 
about % inch wide at the widest part. These divisions are 
trifid above, each sttbdivision nearly 4% inch wide, rounded at 
the tip and mucronulate. Stem leaves few and similar; the 
apex of the follicle straight and not outcurved as in Parryz. 
Other two specimens in the Parish herbarinm froin mountains 
near Palm Springs 1222, and April, 1896, Palm Springs, and 
one in the author’s herbarium from Palm Springs, are the only 
ones in which the flower parts quite coincide with the original 
description. Mr. Parish’s annotations on his own collection 
bear out this view. The flowers of Parzshzz are of a lighter blue 
than those of Parry. 
D. HESPERIUM Gray. When Greene made D. Hanseni a 
species on account of its peculiar seed appendages, he went a 
long way toward simplifying the supposed variability of this 
species. In examining the material under the names of hesper- 
zum and Hansenz in the various herbaria, I am unable to use 
the character of the seed to determine specific differences, as 
only one of all the numerous plants available possess any ripe 
fruit (Agricultural Station, Amador county, 104 Hansen). The 
fruit of this is as described by Greene. To fall back on the 
general vegetative characters without examining the seed is 
scarcely a scientific procedure, and my opinion in this matter 
must be considered merely tentative. The flowers vary so much 
in size in those that otherwise would be classed as Hansenz, that 
