24 PITTONTA. 
SEGREGATES OF VIOLA CANADENSIS. 
To any one actually acquainted, by travel and experience, with 
the topography and climatology of the territory of the United 
States, East and West, it will seem incredible that any one native 
species of violet should really occupy the whole range assigned 
this one by botanical compilers, that is, from the Atlantic to the 
Pacific, and from near Hudson’s Bay to the borders of Mexico. 
More than suspicious that this name as in use covers an aggre- 
gate of several species, I have, at intervals during a dozen years 
past, examined the materials so named in my own and other ex- 
tensive herbaria, but not until recently with results sufficiently 
clear to warrant my attempting a breaking up of the presumed 
aggregate. And even now, while some of the segregates to be 
proposed are clearly enough defined as species, others are less 
so; and my work as a whole is somewhat tentative and pro- 
visional, as indeed all taxonomic work is, and for generations 
yet to come must continue to be. 
In this group, as in others where the typical species is eastern, 
one names southern and far-western segregates without fear of 
creating confusion in nomenclature. The original Viola Cana- 
densis was no southern or far-western plant. Yet in even the 
northeastern States and Canada—the region whence the real 
V. Canadensis was derived—there is some diversity of recogniza- 
ble and definable forms, and the question, to which of these 
does that name really belong, is a perplexing one; perhaps not 
satisfactorily to be answered by any one at this late date. Never- 
theless, as our first task must be that of defining a Ae to bear, 
in our mind, the old name, I shall at the outset give the char- 
acters of the plant which goes by that name in central New 
York, where it appears to be more common than elsewhere as 
far as the collections indicate. 
V. CANADENSIS, Linn.? Stemsa foot high more or less, from 
