BRS a A Ne ek A ah a cies Ll nl) eS ee eA 
' 
1896 | ROSA AMERICANA: 9 
show us glabrous or pubescent leaves, simple or glandular- 
compound leaf-teeth, leaves glandular or non-glandular beneath, 
and the floral organs smooth or hispid-glandular. 
Instead of describing at length all these variations, a work 
reserved perhaps for a monograph, in would seem best to present 
the characters proposed to distinguish R. Nutkana from the neigh- 
boring species. 
I shall compare it first with R. d/anda, which it resembles in 
certain features. The armature, when it is normally developed, 
is sufficient alone to distinguish the two types. Both, as is the 
case habitually among the CiInNAMOME, bear numerous ordi- 
narily setaceous? prickles below, but in R. Vutkana these prickles 
are accompanied by stouter prickles regularly paired on the 
leaves, prickles which are entirely lacking in R. dlanda. The 
paired prickles of R. Nutkana, on the stems especially, are very 
peculiar and very different from the prickles of other American 
species of this section, often being thick, triangular in form, and 
more or less decurrent at base, normally straight and perpen- 
dicular to the axis from which they arise. On the branches these 
paired prickles become less robust, and likewise on the floriferous 
branches. Watson describes the prickles of R. Nutkana as being 
decurved, and Best says the same. I have examined sufficient 
material, both wild and cultivated, to be well assured that the 
prickles belong to the straight type, only exceptionally becoming 
decurved and hooked, as is the case in European and Asiatic 
species with prickles of the straight type. But the decurving is 
simply an accident, and it is this accident which Watson and 
Best saw. Prickles normally curved and hooked apparently 
occur in but one American species of the section CINNAMOMEZ, 
namely, R. Californica. 
The paired pricklesalso distinguish R. Nutkana from R. blanda, 
which is absolutely free from them. But if. for some reason the 
paired prickles disappear from certain parts of the axes, it is 
then necessary to have recourse to other distinctions. Such are 
9In the variety of 2. Nutkana which I formerly called R. Durandii the setaceous 
spines are often replaced by pedicellate glands. 
