STUDIES IN THE CRUCIFER. 135 
repeatedly observed that it has shown a decided preference 
for railway grades and embankments; and along these lines 
it has certainly already passed beyond its natural habitat. 
I suspect that in the Minnesota Valley, where it is said to 
occur along railway tracks, it is a somewhat recent impor- 
tation. 
11. C. Arxansanus. Erysimum Arkansanum, Nutt. in 
T. & G. Fl. i. 95. What might be expected to prove thor- 
oughly good characters are attributed to this southwestern 
plant, in the place where it was first published: and, from 
the careful examination of very numerous specimens gath- 
ered in Missouri, Arkansas, Texas, New Mexico, ete., I find 
that they hold well. As compared with C. asper, the pods 
are,as Nuttall said, tapering at summit, and they are also 
shorter and less spreading. The pubescence of the leaves is 
also remarkably unlike that of C. asper, though Nutiall’s 
description of it as 3-parted is not quite accurate. The hairs | 
are cleft to near the middle, but so appressed as to seem 
3-radiate. The leaves are constantly broader and thinner 
than those of the plant of the far northern habitat. Its 
nearest relative is not C. asper, but 
12. C. ELATUs. Erysimum elatum, Nutt. l.c. Not easily 
distinguishable from the preceding by any difference in the 
pubescence; but foliage narrower, mostly entire; flowers 
much longer, but petals narrower, probably arranged irreg- 
ularly, after the manner of other far-western species. This 
is traceable throughout Oregon and California, in the foot- 
hills of the Cascades and Sierra; though possibly the plant 
ofextreme southern California, with broader greener foliage, 
may be identical with C. Arkansanus. 
13. C. WHEELERI. Erysimum Wheeleri, Rothr.in Wheeler's 
Report, vi. 64. While Mr. Rothrock's description clearly 
enough indicates a rather strongly characterized species, 
