Mr. Schweinitz on the Genus Viola. 49 



bed by authors. But in the execution of this plan, I found 

 myself under the imperious necessity of proposing several 

 new species, as a means of extricating some of those hith- 

 erto received, from the attending difficulties. It has been 

 my endeavour not to do so upon slight grounds, but on the 

 contrary to satisfy myself by reiterated observations, m 

 most instances continued for several years, of the constan- 

 cy of the characters, I have selected. The ample descrip= 

 tions I have given of all, I had an opportunity of studying, 

 in a fresh state, will enable Botanists to test the accuracy 

 of my discriminations. 



The generic description of the genus given by Mr. Nut- 

 tall (p. 147, Vol. I.) is so satisfactory, that it is needless to 

 repeat it. His observation concerning the apetalous flowers 

 of all our species, I entirely accord with, although they ap- 

 pear not so regularly in the Caulescent family, as in the first ; 

 and it is impossible not to coincide with him as to the anom- 

 olous V. concolor, which evidently, together perhaps with 

 some tropical species, ought to be genericaliy separated 

 from the rest. All the other American species form, with 

 those indigenous to northern Europe, (the greater part of 

 which are in my collection,) a most excellent natural genus. 

 There is however no single European species that can 

 with propriety be considered identical with any of ours, un- 

 less it be the V. bicolor and even this differs so much, that 

 I cannot think them the same ; indeed, it is rather remark- 

 able that no European Viola, except very partially the V. 

 odorata, appears to have become naturalized in America, 

 not even the tricolor, so extremely common in Europe in 

 fields of grain. 



The American species, like those of Europe, form two 

 very handsomely distinguished groups or families ; the first 

 altogether Stemlcss (acaules,) and the other producing 

 stems (caulescentes.) With us the number of species in 

 each is nearly equal. The first group contains only four 

 European species— the last from 16 to 17 ; a considera- 

 ble proportion of which are alpine, and belons: to the sub- 

 division, Stipulis pinnarijidis, of which we have but one. I 

 presume some more, and perhaps new ones might be found 

 in our highest northern mountains, or in Labrador. 



Genus Viola. 



Classis V. Pentandria. Ord. I. Monogijnia. 



Vol, v.— No. 1. 7 



