Violets of the Atlantic Coast. 91 



puv})le-flowere(l violets are here maintained as distinct. Pursli 

 and .Schweinitz, two of the earliest authorities in this century, 

 recognized each ten species, Nuttall accepted six, Le Conte thir- 

 teen, and Torrey and Gray six. In the first edition of the INIan- 

 ual, Gra}^ admits hut four species, in the second five, and in the 

 fifth and sixth editions three only. In the most recently pub- 

 lished work, the Synoptical Flora, above referred to, there are 

 included three species and four varieties. It seems obvious that 

 the most logical course of procedure for a conservative botanist 

 is the reduction of all possible forms to the Linn^an species 

 2:)almata, for the differences between pahaata and sagitiata, the 

 validity of both of which is everywhere admitted, are scarcely 

 more than those l^etween any others of this group selected for 

 comparison. 



Synopsis of Species. -^ 



Leaves all pedately divided ; rootstock short and al)nii)tly 



perjiendicular I', pedata. 



Leaves broadly lobed or undivided ; rootstock ascendinu- or 

 horizontal. 



Pleads glabrous or tvitit, verji sliglit pubescence : 



Leaves somewhat pinnately 7-lobed T'. septemloba. 



Leaves deltoid-cordate or panduriform V. dentata. 



Leaves hastate or sa,ti;ittate, basally incised. . . T'. sagiitain. 



Leaves cordate-cucullate V. obliijua. 



Pl/ails pubescent or rnions : 



Leaves pahnately lobed b. pedmaln. 



Leaves ovate or oval T'. ovata. 



Leaves cordate-orlncular T". villosa. 



Viola pedata L., Sp. PI. 933, 1753. f Not of subsequent authors. 



r. pedata bicolor Pursh, tide Raf. in D. C, Prodr. 1 : 291, 1824. 



Viola pedata inornata Greene, Pitt. 3 : 35, 1896. 

 V. pedata of autliors, not of L. 



* In this connection it should be stated that V. pedatificla Don, which 

 is closely related to V. ]>ed(da, is omitted as not belonging strictly to our 

 coast. 



tProf. E. L. Greene has proved that the type of the Linnpean pedcd.a 

 must have been a i)lant of the bicolor variety rather than the mono- 

 colored form which we are accustomed to regard as pedata. This is con- 

 clusively shown by an examination of the plate of Plukenet to which 

 Linnaeus refers. 



