THE TRUE POSITION OF VIOLA TRIPARTITA ELL.
The record of the occurrence of Viola hastata in Florida has al ways been a puzzle to me, for that plant is as typical an Alle ghenian species as our flora affords.
During my field work in the Southern States I have had oppor tunity to study the forms under consideration in their native habi tats and have been led to the following conclusions, namely : That V. tripa•tita is specifically distinct from its relatives, and that it is closely related to V. pubescens and V. scabriuscula and not to V. hastata. It seems strange that Dr. Gray, and even the saga cious Le Conte failed to observe the latter fact, for a casual glance at the foliage and flowers is sufficient to prove this position cor rect, the sepals and petals of V. tripartita being much more similar to those of V. pubescens than to those of V hastata.
The question that naturally arises is : How was V. trIpartita ever confounded with V. hastata? As far as I can see this was brought about through observations on simple-leaved plants of V. tripartita; it is on simple-leaved specimens of this plant that V. hastata is admitted to the flora of Florida. The leaf form of
the latter species is so distinct and unique that it need not be fur ther considered in this connection. On the other hand the leaf form of V. tripartita in its simple state, which is quite frequent, closely resembles that of V. pubcsans but differs in size, shape, proportionate width and length and the toothing. It may be of interest to note that I have seen the type of V tripartila on several occasions and have collected specimens almost iden tical with it on Stone Mountain, which is no great distance from the original locality, Athens, Georgia. I have also received excellent and typical specimens from Mr. A. M. Huger, collected in Polk County, North Carolina, and a series of specimens show ing all degrees of gradation from the simple-leaved state to the trifoliolate leaf, from Mr. E. R. Memminger, who independently came to the conclusion that the affinities of V. tnpartita are with V.pubescens and not with V. hastata.
I append a description taken from living plants.