330 BOTANICAL GAZETTE [NOVEMBER 
CLASs 7. Orbiculares. Corolla yellow ; cleistogenes deflexed 
31. V. rotundifolia 
VIOLA PEDATA L. Sp. Pl. 933. 1753. 
V. digitata Pursh, Fl. Am. Sept. 1:171.. 1812. 
V. pedata bicolor Bursh, fide Raf. in DC. Prodr.1: 291. 1824. 
V. pedata inornata Greene, Pitt. 3:35. 1806. 
This, the common bird’s-foot violet, appears, as is well 
known, in two forms, in one of which the petals are pale blue 
and concolorous, in the other the uppermost petals deep velvety- 
purple. It is the latter form, so long known as the variety 
bicolor, which is figured by Plukenet, and on which Linnaeus 
based his type. Between these two extremes one finds in Mary- 
land and Virginia every possible variation. Professor Greene 
considered the pedata of New England, which is usually the 
concolorous form, distinct from this variable southern plant, 
and bestowed on it the varietal name inornata. I have not 
been able to find characters upon which to base such a separa- 
tion. 
VioLa BERNARDI Greene, Pitt. 3: 260. 1898. As synonym. 
V. pedatifida var. Bernardi Greene, |. c. 
Professor Greene applies this name to a violet very familiar 
to me from a specimen in my own herbarium, collected in 
Wisconsin by Professor S. M. Tracy some years ago, which has 
since passed into the possession of the National Herbarium. 
The plant was so remarkable that Dr. Britton originally considered 
it distinct, and gave it a manuscript name. In the states of 
Michigan, Wisconsin, and northern Illinois it seems to be the 
prevailing form, and apparently entirely replaces V. palmata. 
Recently specimens have come to me from Chicago collected by 
Dr. Moffatt, which in addition to Dr. Greene’s material show the — 
plant to be a species intermediate between the Pepat# and the 
HETEROPHYLL#, having the foliage and aspect of the formet, 
together with the rootstock and variable leaf-contour of the 
latter. I have been unable to discover cleistogamous flowers 
upon the material thus far examined. 
