MUTATIONS IN VIOLA. 183 
and figured, but the petals are so evidently notched that this 
must be accepted for one mark of the form; this notwithstand- 
ing that there are occasional hints of the emarginate in those of 
V. pedatifida. 
V. BERNARDI, Greene, Pitt. iii. 260, also partly (and especially 
as to probable origin) Pitt. v. 123, and partly that of Pollard in 
Britton, Man. 635. Stout and low, the whole plant at petalifer- 
ous flowering only 4 or 5 inches high, the hirtellous petioles no 
more than twice the length of the leaves, these flabelliform, often 
broadly and roundedly so, yet only subtruncate at base, never 
cordate, but the middle basal portion tapering abruptly to the 
petiole: petaliferous flowers borne barely above the foliage, large, 
deep purple as to the petals; sepals lance oval, obtuse, somewhat 
serrulately ciliolate. Autumnal state not taller than the vernal, 
glabrate, copiously fructiferous from apetalous flowers, the cap- 
sules borne quite above the foliage and large, oblong, fully 4 inch, 
obtuse, of more than twice the length of the small narrow long- 
auricled sepals. Leaves of all stages cleft to near the middle 
into 9 to 12 oblong straightish (not falcate) obtuse segments. 
This is the first proper diagnosis that has been given of what 
I had in mind as V, Bernardi. Much of what I and others have, 
in the last seven or eight years, referred to it I now see must be 
excluded. The actual type specimen of H Bernard: is a petal- 
iferous one in my herbarium collected by me at Albion, Wis., 
May, 1866. Doubtless the very same, specifically, is one from 
Plattsville, in the same part of Wisconsin, by S. M. Tracy, 1 
Noy., 1887, also in U. S. Herb., this my type for the autumnal 
state ; but there is a better sheet of precisely the same from Riley 
Co., Kansas, 9 Sept., 1895, by J. B. Norton. 
These autumnal specimens present points of contrast with 
those of V. pedatifida which must here be indicated. A perfect 
type of the last isin my herbarium, collected by myself in a 
typical prairie locality near Sandoval, Illinois, 1898, is just a foot 
high, its fruiting peduncles only 5 to 6 inches, therefore borne 
distinctly and far below the leaves. Another sheet by me, taken 
at Prairie du Chien, Wis., the same year, is 10 inches high, with 
peduncles of only 2 to 3 inches, or of less than one-third the 
