202 Pollard Two New Violets. 



The publication of, this species is the result of a critical study of the 

 violets belonging to the hastata group in the South, wherein my own 

 field observations have been supplemented by information kindly given 

 me by other collectors. For many years Elliott's V. tripartita was re 

 garded as a variety of V. hastata, although the most superficial examina 

 tion of the rootstock in the two species ought to have dispelled such an 

 opinion. Yet even after Dr. Small had demonstrated this character it 

 was noticeable to southern botanists at least that certain forms remained 

 which could not well be disposed of in connection with either of these 

 species. More recently Mr. Roland M. Harper* attempted to solve the 

 difficulty by the reinstatement of Gingins' V. hastata glaberrima^ , which 

 he treated as an entire leaved variety of tripartita, remarking in this 

 connection: "It differs from typical tripartita in having leaves all un 

 divided and glabrous, but is in other respects very similar. It seems to 

 extend farther south than the type, and is the plant which was taken 

 for V. hastata in Florida." But Mr. Harper has here fallen into the 

 very natural error of placing in a single category all the plants with un 

 divided leaves, regardless of range, and his statement in the concluding 

 portion of above quoted paragraph is also a trifle too broad. While much 

 of the material referred by Dr. Chapman to V. hastata glaberrima has 

 nothing in common with hastata, the latter species does, notwithstanding, 

 occur in Florida, as is abundantly proven by excellent specimens in the 

 Biltmore Herbarium, collected in Liberty County. A mere glance at 

 the whitish, succulent, tuberous rootstock is sufficient to establish the 

 identity of this species. 



Let us now examine the range of the dubious forms under considera 

 tion. The type of V. tripartita came from Athens, Georgia, and the 

 species has been collected there by many different persons since Elliott's 

 day, so that we have no difficulty in establishing a positive diagnosis of 

 its characters. North of this point it ranges through the mountains of 

 North and South Carolina and eastern Tennessee, being common near 

 Biltmore, N. C., where I have carefully studied it. Throughout this 

 range, and nearly always associated with the typical form, as shown by 

 notes on collector's labels, is a plant of almost identical appearance, ex 

 cept that the leaves are undivided and have a tendency to become glabrous 

 with age. This is the true glaberrima, which may well stand as a variety 

 of tripartita, where it has been placed by Mr. Harper. 



If now we examine the collections from south of Athens, we find an 

 entire-leaved yellow-flowered violet represented in abundance, but no 

 tripartita. It is this plant which was familiar to Dr. Chapman in Florida* 

 and which Mr. Harper has confused with the more northerly form. V. 

 tripartita and its variety are relatively tall, coarse plants, even at flower 

 ing time, having flowers with the corolla distinctly veined, and with rather 

 broad sepals. This violet is very slender, of remarkably uniform size, 

 with smaller flowers, quite immaculate petals, and leaves strongly sug- 



*Bull. Torr. Club. 27: 337. 1900. 

 J-Gingins in D. C. Prodr. 1: 300. 1824. 



