Mr. Schweinitz on the Genus Viola. 53 



Floribus majusculis, violascentibus et purpureis, contor- 

 tis ; petalis lateralibus maxime albo-barbatis, ceteris gla- 

 bris ; inferiori ovato, postice naviformi, manifestim carina- 

 to, antice rotundato, venis purpureis ornato ; omnibus fun- 

 do albescentibus. Cahjcis laciniis ovato-lanceolatis, glabris, 

 raargine albescenti ciliato. Stigmate capitato, recurvo, ros- 

 trato, marginato, depresso. Ardheris in fundo reconditis. 



Scapis longitudine foliorum, interdum glaberrirais. Sti- 

 pulis minutissimis, oppositis, in medio scapo, aut inferius 

 sitis, linearibus, ciliatis. 



Capsulis glabris, calyce brevioribus. 



Radice obliqua crassiuscula, caespites pauci-foliosas ef- 

 formante. 



This is by far the most variable species, and consequent- 

 ly the most perplexing. Scarcely two individuals are per- 

 fectly alike, and I can hardly figure to myself any possible 

 formation of palmately lobed leaves, which I have not met 

 with. Every different situation, as to shade, soil and per- 

 haps even weather, appears to alter the shape of the leaves, 

 and I think it useless to attempt to reduce these ever vary- 

 ing forms to definite varieties, as they appear multiplied al- 

 most without end. The ii^iddle lobe is however constantly 

 the largest ; in some varieties it assumes a long lanceolate, in 

 others a broad ovate shape, with every possible intermedi- 

 ate one. The lower segments or lobes towards the base 

 again, are always the most diversified. In some they are 

 broadly truncate, in others almost linear and elongate. 

 The degrees of pubescence of leaves and petioles, between 

 an incane villosity and almost smooth, are equally variable. 

 The varieties occurring in rich shaded woodland, are par- 

 ticularly remarkable. One I have repeatedly met with, is 

 exceeding tall and slender, pubescent in the petioles and 

 scape, nearly smooth in the leaves, of which it generally 

 has one cordate, one reniform and one very regularly five 

 lobed. The variety dilatata of Elliott might be called the 

 type of the species. A remarkable circumstance is, that 

 the earliest flowering specimens, appearing in March, al- 

 ways put forth one apparently naked flowering scape, with 

 a single undeveloped leaf close to it, and another proceed- 

 ing from a horizontal radicle or stolo at some distance from 

 the flower. The best general characteristic to distinguish 

 this, at first blush, from the next, appears to me to be this, 



