ENGELMANN—NORTH AM. SPECIES OF JUNCUS. 493 
bus binis ternis rarius singulis; sepalis stamina fere duplo su- 
perantibus; antheris oblongis filamento bis brevioribus; stylo 
brevissimo cum stigmatibus brevibus incluso; capsula caly- 
cem quante vel vix excedente obtusa brevissime mucro- 
nulata. 
Var. y. uniflorus ; planta minima (4-1-pollicaris) ; floribus 
bracteis 2 suffultis singulis plerumque 2-meribus (sepalis 4, 
staminibus stigmatibus carpellisque binis).— J. saginoides, 
p. 436. 
California, from the coast to the mountains; var. a. Yose- 
mite Valley, alt. 4,000 feet, Bolander, Hb. n. 30; fl. July; 
var. 6. Ukiah, Mendocino county, the same, Hb. n. 31, #1. 
May, also “ Fort Bragg, near the coast” (1-3-flowered) ; var. 
y: Sierra Nevada, among mosses, //illebrand; upper Tuolum- 
ne River, Bolander, and in the lowlands, Anderson Valley, 
the same, Hb. n. 82; fl. April and May. 
A curious and suggestive little plant, which must consider- 
ably undermine our confidence in certain characters, con- 
sidered of specific value, already shaken by the variations of 
other species from the same wonderful country ; it proves that 
the singleness or plurality of flowers on the peduncle, the num- 
ber of their parts, and, if my view is correct, even the propor- 
tion of stamens and styles, are not sufficient to establish spe- 
cific distinction. The first points are established beyond a 
doubt by some of Mr. Bolander’s specimens from the moun- 
tain region, intermediate between g and y with one or two 
flowers, and often with a dimerous and trimerous one in the 
same inflorescence. Var. @ may be considered a distinct 
species by those that hold its differential characters to be of 
paramount importance, but the similarity of the whole ap- 
pearance of the plants and of most of their parts, and, above 
all, the absolute identity of the well-marked seeds, convince 
me that it must be united with the others, and that eventu- 
ally intermediate forms will dispel all doubts. 
Only the small dimerous form was known to me when the 
first part of this paper went to press, and was then considered 
as the type of a distinct subgenus, Juncellus, allied through 
its single-flowered stems to Lostkovia, and distinguished by 
its dimerism from any other known Juncus (see pp. 426, 428 
& 436). Mr. Bolander, however, has since discovered other 
forms of this plant which bear trimerous flowers, thus assimi- 
lating it to the ordinary form of Junci and more particularly 
to the European J. capitatus, and destroying the subgenus 
Juncellus. Iam now convinced that it must be placed with 
its Kuropean ally near J. marginatus, in the section Grami- 
net, the dimerous variety constituting an anomaly not other- 
wise observed in this genus, but again found among the 
allied Restiacee and Lriocaulonee, where dimerism and 
