83 
throughout; empty glumes 1} to 2 lines long, the lower slightly longer, oblong-lance- 
olate, acute, whitish, translucent, nearly smooth except on midnerve; floral glume 
thin, one-fourth shorter than the empty ones, oblong, obtuse, the 4 nerves terminating 
in as many teeth at the apex; a strong awn from near the base to a little beyond the 
glume; hairs of callus very few and short (} line), those of the rudiment twice as 
long, or half as long as the glume; palet about as long as its glume.—Montana, British 
America, Washington, Oregon and California, 
21. C. Tweedyi Scribn. Culmsrobust, thick, smooth, 3 to 3} feet bigh ; leaves flat, 
unusually broad, those at the base 3 to 5 lines wide and 6 to 3 inches long, those of 
the culm 1 to 3 inches long; ligule 1 or 2 lines long, decurrent; sheaths loose, striate, 
smooth; panicle spikelike, about 4 inches long, thick, densely flowered; empty 
glumes 3 lines long, thinnish, lanceolate, acute; floral glume slightly shorter, lance- 
olate, finely scabrescent. acutely 2-toothed, with a stout, twisted, and bent awn 
from the lower third to one-third or one-half beyond the glumes; palet as long as 
its glume, 2-toothed; hairs of callus very short and scanty, those of the rudiment 
numerous and half as long as the floral glume.—Cascade Mountains, Washington 
(Tweedy, GR. Vasey). 
Remarkable for its robust size, broad leaves, and thick panicle. Near C. sylvatica. 
22. C. sylvatica DC., var. AMERICANA. Culms tufted, erect, rigid, 1 to 2 feet high; 
leaves rather numerous and crowded near the base, rigid, the radical ones more than 
half the length of culin,those of the upper culm 3 to 6 inches long, attenuate-pointed, 
very scabrous, thick, and more or less involute; ligule 1 line long, truncate often 
lacerate; sheaths rough, striate; panicle spikelike, strict, sometimes loose, 8 to 4 
inches long, 4 inch wide or more, dense, generally purplish; rays mostly in fives, an 
inch long or less, appressed, and like the rachis very rough; spikelets 3 to 3} lines 
long, on short roughish pedicels; empty glumes ovate-lanceolate, very acute, scab- 
rous, the upper 3-nerved; floral glume slightly shorter than the empty ones, 
very acute or acuminate, with 4 setaceous teeth at apex, awn near the base, twisted 
below, bent at the middle,and exserted one-half or quite as long as the glume; palet 
hyaline, narrow, 2-toothed ; hairs of the callus scanty, short (about 4 as long as the 
flowering glume), those of the rudiment $ or ¢ as long as the flowering glume.—High 
plains and mountains, British America to Colorado, also Washington, Oregon and 
northern California, 
This differs considerably from the European form. 
Var. LONGIFOLIA. A remarkable form from the Mattole district, California. Culms 
densely tufted ; the radical leaves as long as the culm, almost filiform and setaceously 
pointed, empty glumes narrow, 4 lines long, acuminate; hairs of the callus more 
conspicuous, one-third as long as the glume. -A well-marked variety. (No, 6470 Bo- 
lander). 
Var. PURPURASCENS Thurb. is a more slender, less rigid form.—Cal. (No. 5071 Bo- 
lander). 
23. C. purpurascens R. Br. (Hooker FI. Bor. Am, u, p. 240.) Culms tufted, firm, 
not rigid, 2 to 24 feet high, leafy; leaves flat, 6 to 8 inches long, 2 to 4 lines wide, 
somewhat scabrous, tapering to an attenuated point; sheaths shorter than the in- 
ternodes, smooth; ligule 2 lines long, lacerate; panicle spikelike, 3 to 34 inches long, 
one-half inch wide, purple; empty glumes 2 to 24 lines long, equal, broadly lanceo- 
late, smoothish, hardly scabrous, acute; floral glume one-quarter shorter, oblong, 
obtuse at apex, smooth; awn at lower third and extending scarcely beyond the empty 
glumes; palet equaling its glume, obstusely 2-toothed ; hairs of the callus very short 
and scanty, those of the rudiment unequal, about half as long as the glume, 
This description is from specimens collected in the Cascade Mountains of Washing- 
ton by G. R. Vasey, which agrees well with European specimens, although differing 
somewhat from the description in Hooker’s Fl. Bor. Am. It occurs also in British 
America and in Sitka. It is not C. sylvatica var. purpurascens of Thurber, 
24. C. rubescens Buck? Prof. Scribner has identified as this species specimens 
