1907] Brainerd,— Older Types of North American Violets 97 
2. This descriptive term better characterizes the plant of the low- 
lands than that of the uplands. 'The pubescence of the former is, 
indeed, rather short for the precise use of the word ‘villosa’; but it is 
dense and soft, that of the petioles and peduncles being, as Prof. 
Greene puts it, “almost plushy." Mr. Pollard also says of it in Small’s 
Southern Flora, “petioles usually densely villous." But of the upland 
plant Prof. Greene remarks, “the accepted V. villosa is not villous; 
it is rather stiffly hirsutulous." 
3. The opinions of Elliott and of LeConte are entitled to great 
weight regarding the point in question. Through the kindness of 
Prof. Paul M. Rea, the Curator of the Charlestown Museum, I had 
the privilege of examining the specimens of Viola in the Elliott her- 
barium, and found the plant that he had labelled V. villosa Walt. to 
be the species of the coastal plains. LeConte, than whom no botanist 
in the first half of the last century more diligently studied and more 
accurately discriminated the perplexing forms of Viola in the Atlantic 
States, is most emphatic in applying Walter’s name to the plant of 
the coastal region, and not to the plant of the northern uplands. 
“When V. villosa is misunderstood,” he justly remarks, alluding 
to Nuttall and Torrey, “it is simply by authors who have never seen 
it." LeConte’s description of the two species as well as his accurate 
figures, now in the possession of Prof. Greene, confirm these conclu- 
sions. This interpretation of V. villosa was adopted in Torrey and 
Gray’s No. Am. Flora, in 1838, and is the one maintained in all the . 
many editions of Eaton’s Botany published after 1830. 
The true V. villosa Walt. has the unique habit of bearing apetalous 
flowers and fruit in the winter and early spring, before bearing petal- 
iferous flowers. Much of the autumn foliage remains green through 
the winter, and specimens that I collected the last week in March had 
abundance of ripe seed but no petaliferous flowers. In no. 31 of 
Greene and Pollard's distribution of No. Am. Violaceae are to be seen 
both flowering and fruiting specimens of this species, collected March 
30 and April 19; but the fruit is from cleistogamous flowers and was 
evidently obtained on the earlier date, the flowers on the later. 
The other species, that is currently passing as V. villosa, will need 
1 It should be understood that both these authors are here speaking of V. Carolina, 
apparently not suspecting it to be Walter’s V. villosa. 
2“ Cum V. villosa confusa, auctoribus duntaxat quibus nunquam visa." Ann, N. Y. 
Lyceum, ii. 144, 
