14 BOTANICAL GAZETTE [JULY 
it prudent to obtain additional information before uniting them, 
Therefore in the dichotomous table at the end of this paper I 
have separated RK. Arkansana as a distinct species, for the purpose 
of facilitating investigation. It is especially necessary to examine 
the mode of vegetation of R. Arkansana, and to discover the 
modifications involved in dwarf and giant forms. The cause of 
the dwarfing, which is quite frequent, should be investigated. 
Perhaps the frequent fires which ravage the prairie region where 
KR. Arkansana appears to occur may have some connection with 
the dwarfing. It is important to know whether R. dlanda, more 
or less typical, does not occur quite frequently associated with &. 
Arkansana, and whether the latter does not gradually pass into 
the former by a series of intermediate states. 
I have specimens of typical R. Arkansana from Minnesota, 
Nebraska, Kansas, Colorado, and Manitoba, and from the Sas- 
katchewan. I have described the plant from the last station 
under the name &. dlanda var. setigera. 
Finally, if R. Arkansana is found to be a variety of R. dlanda, 
the area of distribution of the latter will be extended chiefly 
westward. 
Rosa pisocarpa A. Gray. 
In 1876 (Prodromus) 1 expressed doubts as to the autonomy 
of X. pisocarpa A. Gray. These doubts have been removed by — 
subsequent study of abundant material received from America, 4 
and from cultures which I have made. Before entering into the _ 
discussion, I wish to remark that Nuttall, who first had recognized 
the character of R. gymnocarpa and R. foliolosa, and likewise of 
R. Nutkana, which he had designated in his herbarium as 2. mega- 
carpa, had taken R. pisocarpa for a new species, to which he gave 
the name R. arguta MS. If Torrey and Gray had retained these 
two names proposed by Nuttall, there would have been two more 
species to his credit. 4 
In its habitual form, such as Asa Gray described and figured . 
in the Botanical Magazine (pl. 685 7), R. pisocarpa cannot be con- 
founded with any other species. Its prickles regularly paired, 
and its leaflets more or less conspicuously rounded at base, dis- 
