1903.] NATURAL SCIENCES OF PHILADELPHIA. 689 



Viola rostrata Pursh is common in the Alleghanies and comes south- 

 ward to Bucks county. Muhlenberg's name is unrecognizable until 

 clearly defined by Pursh; his specimens, however, show that both 

 writers had the same plant in view. 



Viola canadensis Linn, is also abundant in our mountains, and there 

 is a specimen in the Academy herbarium marked Sellersville, Bucks 

 county, the farthest south that I have heard of its occurrence. 



30. Viola rafinesquii Greene. 



"Viola hicolor" and '' arvensis" of several American authors (nee European 

 authors). 

 VioJa tenella Muhlenberg, 1813, Catalogue, p. 26 (nee Poiret, ISIO). 

 Viola rafinesquii Greene, 1S99, Pittonia, IV, p. 9. 



Range. — Along the Susquehanna and Delaware rivers, and abundant 

 at various localities in southern New Jersey outside the pine barrens. 



Habitat. — Low sandy ground. 



Description. — Flowering plants, Nottingham, Chester county, 

 Pennsylvania, near Octoraro creek. May 6, 1902. Benjamin H. Smith. 

 No. 5,171, Herb., W. S. Stems glabrous, 100-150 mm. high, often 

 branching close to the base; stipules conspicuous, deeplj' cut, lobed; 

 leaves 20 mm., linear-lanceolate or spatulate, the lower nearly orbicu- 

 lar (10 X 10 mm.), contracted into margined petioles, 8-10 mm. in 

 length, obscurely toothed, glabrous. Flowers several in the upper 

 axils on pedicels 25 mm. long, yellowish with a blue tinge, lower petal 

 lined with black, 15-18 mm. wide, sepals rather broadly lanceolate- 

 acute. 



Plants in full fruit May 30, same size as above. No further growth 

 takes place, and the plants soon wither up. 



Nomenclature and Synonymy. 



Difficult as it is to determine tlic number of distinct forms that exist 

 among our violets, it is a still harder prol)lem to ascertain what names 

 they should be known by. Some 160 specific or varietal names have 

 been proposed for violets of eastern North America, and as many 

 of these are very meagrely descri])cd, it is well-nigh impossible to be 

 sure what forms the authors had in hand. 



In the preceding pages I have entered in the synonymy only such 

 names as seem without reasonalolc doul)t to refer to the speci(>s in ques- 

 tion, and have discussed matters of nomenclature only so far as abso- 

 lutely essential to the determination of the proper name to use in each 

 case. In order, however, to satisfactoril.y cover the synonym}- of the 

 genus, I have prepared a list of the names known to me which have 

 been applied to eastern North American violets — both specific and 

 44 



