1896 | ROSE AMERICANLE 29 
unarmed. In this case it is easy to distinguish the two species ; 
but if A. dlanda increases its armature by extending it to the 
middle regions of the bush, and even to certain floriferous 
branchlets ; or one finds &. Arkansana with axes completely seti- 
gerous, then recourse must be had to other characters. Those 
taken from the leaves are most practical. The leaves of X. 
acicularis are oval, rounded at base, almost always glandular 
beneath, with open teeth, margins glandular-denticulate or pro- 
vided with glands; while those of R. dlanda and KR. Arkansana 
are oboval, relatively narrower, quite strongly attenuate or cune- 
ate at base, very rarely glandular beneath, with teeth almost 
always perfectly simple and directed towards the apex. The 
form of the stipules is a little different in R. acicularis. The 
inflorescences are much more frequently one-flowered in &. act- 
cularis than in R. blanda and R. Arkansana. The statistics that 
I have obtained show that the one-flowered inflorescences are to 
the many-flowered inflorescences as 3.4 to I in R. acicularis; 
while in &. dlanda they are as I to 1.4, and in R. Arkansana as 1 
to 1.6. Lastly, the fructiferous pedicels of R. actcularis are more 
slender, less rigid, and often have a tendency to become incurved, 
instead of remaining straight. 
As to its geographical range the American &. acicularis occu- 
pies a very considerable area. In latitude it extends from the 
neighborhood of the polar circle to 38° in the Rocky mountains 
of Colorado. In longitude it embraces almost all of the conti- 
nent within the Dominion. To the south it has been observed 
in Michigan, Wisconsin on the borders of Lake Michigan, and 
in Minnesota on the borders of Lake Superior. I have received 
specimens obtained by Messrs. Greene and Kelsey from Helena, 
Montana, and by Mr. Coulter from the Teton region in Idaho. 
It is likely that it exists here and there in the Rocky mountains 
from the borders of Canada to Colorado, where it does not 
appear to be very rare. As yet there is no indication of it in 
Oregon and Washington, or in Vancouver Island. 
In the Old World R. acicularis extends perhaps farther north, 
but it extends less towards the south than in America. 
