160 Rhodora AUGUST 
“banks of the Missouri,” which is the type location of Pursh’s species. 
Besides the discrepancy of range, this Rocky Mountain species differs 
from L. decaphyllus as characterized by Pursh, in having elliptic- 
lanceolate instead of oblong-elliptic leaflets, and in very rarely pos- 
sessing five pairs of leaflets. Although Pursh described his species as 
having only 3—4 flowers, the authors feel, on account of the other 
characters, — the semi-sagittate linear stipules and the five pairs of 
oblong-elliptic mucronate leaflets, — and on account of the type local- 
ity, that L. decaphyllus Pursh was in all probability identical with the 
plant above described as L. venosus var. intonsus. It may be noted 
that Hooker ! made a similar interpretation of Pursh’s species, although 
he failed to recognize its specific identity with the eastern form of 
L. venosus. 
The only other name which has been used for the Rocky Mountain 
species which we have deprived of the name L. decaphyllus, is L. poly- 
morphus Nutt.2 In the original publication of this name, after the 
technical description, Nuttall adds, “Has. On the grassy alluvial 
plains of the Missouri, from its confluence to its sources? Flowers as 
large as those of Pisum maritimum, and of a fine purple, variable 
however in size as well as the leaves, hence it appears to be L. deca- 
phyllus Pu 2. p. 471, and Vicia stipulacea of the same, 2. p. 739. as 
both these specific names are inexpressive and deceptive, I could not 
in candour do otherwise than reject them.” Now, we have already 
shown that Lathyrus decaphyllus Pursh, and Vicia stipulacea Pursh 
are not identical. Excluding, then, Nuttall’s synonym, L. decaphyl- 
lus, his description of L. polymorphus, in greater part at least, is applic- 
able to Vicia stipulacea of Pursh, and we would, accordingly relegate 
it to the synonymy of the latter species, since we cannot reject an 
older specific name as “inexpressive and deceptive.” As will appear 
below, Pursh’s Vicia stipulacea is not the Rocky Mountain species 
now under discussion. It is possible that even more than two species 
entered into Nuttall’s conception of L. polymorphus, but if so, we are 
unable to identify them, and the range given seems to preclude entirely 
the application of this name to the Rocky Mountain species discussed 
above. We are, therefore constrained to describe this well known 
plant as a new species. 
LaTHYRUS eucosmus, n. sp. — L. polymorphus of Torrey and Gray 
in part, and of later American authors, not of Nuttall; L. decaphyllus 
1 Hooker, W. J., Fl. Bor.-Am. i. 159-60 (1839). 
3 Nuttall, T., Genera of N. Am. PI. ii. 96-7 (1818). 
