THE GENUS VIOLA IN MINNESOTA. 129 
for the last species. There are two sheets of petaliferous- 
flowering specimens by Sheldon, the one from Two Harbors, 
the other from Tower, both in St. Louis Co., west of Lake 
Superior; and there isa good sheet of older specimens, taken 
in August, by Sandberg, at Crow Wing. Another and a 
fine specimen in apetalous flowering is by the collector last 
named, who obtained it at Carlton, in July, 1891. This 
specimen is mounted on a sheet with one of V. renifolia. 
Also Roberts’ specimen from Black Point, 24 Aug. 1879, 
labelled by him “V. rotundifolia?,” and which seems to 
form the chief guarantee of the occurrence of V. rotundifolia 
in Minnesota, is plainly V. Brainerdii. 
22. V. RUGULOSA, Greene, Pitt. v. 24, 25. I founded this 
species on a single sheet existing in my own herbarium, 
which specimens. bear so strong a general resemblance 
to V. pubescens rather than to V. Canadensis that I could not 
doubt about its being distinct from the latter. There are 
some thirty sheets of the new segregate in the Minnesota col- 
lection, the greater proportion of which were found labeled 
V. Canadensis, while one has been labeled V. striata. But 
to this “ V. Canadensis” cover some one had transferred per- 
haps recently, four sheets of V. pubescens; this movement 
attesting again, that V. rugulosa and V. pubescens often so 
closely resemble each other that, without knowing the color 
of the flowers, they may get confused in the herbarium. 
While there is no sheet in the Minnesota collection that 
exactly matches my type of V. rugulosa as regards copious 
leafiness, strictness of habit and pronouncedly rugose leaf- 
surface, yet a near approach to it occurs in one from Lake 
Minnetonka, June, 1890, by Geo. B. Aiton. The Sandberg 
sheet the label of which exactly corresponds to mine bears 
Specimens younger than mine, with fewer leaves, the texture 
thinner, the unevenness less pronounced. But the collec- 
