NEW OR NOTEWORTHY VIOLETS. 91 
of V. sagittata in outline, but the body of the leaf above the 
sagittate and incised base coarsely few-toothed: sepals lance- 
oblong, obtuse, strongly ciliate even to the short and rounded 
auricles: corollas more than an inch broad in the larger plants, 
not deep violet, petals broad, subequal, the laterals densely hir- 
sute at about the base of the blade: apetalous flowers almost or 
quite hypogeous. 
The specimens of which the above is a description were gath- 
ered by myself, in an open and rather dry piece of gently sloping 
woodland about half way up Maryland Heights, near Harper’s 
Ferry, 14 May, 1898. There was no other violet growing with 
it, or even in the near vicinity of it; and the about three 
conditions of it, as to cut of leaf, are all so dissimilar that, were 
such a character of any worth, three species might have been 
recognized here, instead of one. Yet, from a diligent study of 
the plants as they grew, supplemented by a critical examination 
and comparison of the carefully prepared specimens in the dry, 
I am of the opinion that such should not be named even as 
varieties. The plants are precisely one thing, in all but the 
fact that some specimens have uncut strongly cucullate leaves, 
others broad deeply palmatifid ones, still others an exactly sagit- 
tate leaf-outline. The one thing of which I am confident is, 
that this mountain woodland plant is perfectly distinct from a 
related woodland violet common enough in lowland Maryland 
and Virginia, and with which I am very familiar. That my 
V. variabilis has not one or more names already, I dare not say. 
But I hold that it is far better, in case like this, where I have 
one species under forms enough to make three, to give ita new. 
name, for purposes of observation and discussion, than to apply 
to it by guess merely, any old name which might or might not 
belong to it. 
The palmatifid phase of V. variabilis is exceedingly similar to 
Mr. Pollard’s recently proposed V. Angelle, indeed so much 
like it that I have little if any doubt they are one, specifically. 
But then again, the uncut and cucullate-leaved form is quite as 
like the type specimens of my own V. populifolia. I can not 
