136 
1” wide, slightly winged or merely margined, terminated by the 
solitary spathe, or with two spathes geminate within the enclosing 
outer bracts. Bracts often conspicuously purplish, very unequal, 
the outer one commonly twice the length of the inner, sometimes 
rigidly prolonged to four times its length, becoming over two 
inches long, sometimes minutely papillose or even papillose-puber- 
ulent; inner bract often appearing gibbous in the spathe, 6-12” 
long, attenuate or acute. Flowers variable, sometimes very, deli- 
cate on slender curved pedicels, often large with the pedicels strict 
and erect; divisions of perianth sometimes 6’ long, the broader 
series 214’ wide, minutely pubescent on outer surface; floral 
scales either clear white or brownish-tinged. Capsules 1-9, glo- 
bose, often larger than in graminoides, on straighter, less exserted 
pedicels, seeds often larger than in gramuinoides, .o4~.00 in. long, 
mostly obliquely obovate-oblong, often angled, brownish, smooth 
or with coarse shallow pitting. (Plate 265.) 
In damp or dry soil, sometimes on sterile hills and in dry up- 
land pastures. 
From Newfoundland to Saskatchewan, south to Connecticut 
and New York and along the mountains to North Carolina, in the 
interior to Kentucky, Missouri and Kansas. 
It would appear from the foregoing description of S. angusti- 
folium that the species is subject to not a few rather striking vari- 
ations. I have little doubt, however, that my description em- 
braces rather a group of closely related plants than the mere range 
of variation shown by a single one. In the material at command, 
however, I cannot find any certain warrant for the subdivision of 
the group, nor can I fully assure myselfas to the value of any one 
of the apparently several component forms. It is probable, in- 
deed, that study of these plants in life shall have to lead the way 
to the correct understanding of their relationships. : 
The true S. angustifolium would appear to range from — 
Newfoundland far westward through Canada and southward 
through the Alleghenies to North Carolina, finding in Connecti- 
cut and New York its southern limit near the coast. Specimens 
agree generally in drying blackish like S. graminoides, and in hav- 
ing the edges of stem and leaves decidedly rough-serrulate. At 
the southern Jimit of its coastwise range this form seems to show — 
a closer affinity to S. graminoides, and it is in plants from this 
region that a short lateral branch is occasionally found. I< is 
quite possible, therefore, that plants practically intermediate be- 
