NEW PLANTS FROM NORTH DAKOTA gI 
views by observation of the similarity of manifestations under 
another latitude. I found that the North Dakota plants have 
some characters in common that distinguish them from their 
southern relatives, and therefore I will in the first place attempt 
to outline those general characters as they present themselves 
within this state as follows: 
Stems, especially their upper part, pubescent with white 
shaggy hairs, 1-5 dm. high, single or several, erect or ascending 
from a large, somewhat woody tuber and bearing numerous or 
comparatively few leaves. Radical leaves long, lanceolate, pro- 
tractedly tapering into very long petioles. The lower stem leaves 
are lance-oblong, tapering into petioles of very variable length. 
Upwardly the leaves becoming narrower and shorter and at last 
bract-like. The leaves are arranged on the stem in two series, 
and they are usually pubescent, sometimes glabrate, but never 
perfectly glabrous. Heads sessile to long-peduncled, of variable 
size, 1 to 12 in a short raceme, but occasionally 30 or more in a 
more or less dense spike or thyrsus. Bracts in 4-7 series, green with 
purple, scarious, erose margins, the outer orbicular, the middle 
rows broadly spatulate, the inner oblong. 
The Rocky Mountain forms as described by Prof. Aven 
Nelson (Liatris ligulistylis) are single-stemmed with glabrous 
leaves, else they appear in general characters to be near relatives 
of our plants. The North Dakota plant—as learned from Mr. 
Deam’s material—differs considerably more from its southern 
relatives, principally in its shorter racemes and shorter involucral 
bracts and in its smaller size, the southern plants having many- 
headed spikes, often several dm. long, and the involucral bracts 
longer, sometimes pointed, more loosely imbricated, often so as to 
make them appear sub-squarrose. They are extremely beautiful 
and striking (one of them looking rather strange with its drooping 
heads). 
When considering the scariosa group in its variety of forms, 
one would feel tempted to compare it with the genus Hieracium 
of the Old World, but the differential characters of the latter 
seem to be easier to systematize. The following suggestion of a 
key for the North Dakota group will be practically useful, though 
the multitude of intermediate forms forbids the application of 
the proposed names as indicating species and causes a great deal 
of hesitation even in using them as variety names: 
