226 Rhodora [Novkmbfr 



A NEW VIOLET FROM NEW ENGLAND. 

 Homer D. House. 



(in;Ue59.) 



Although the described species of Violets have increased greatly 

 within the past decade, it is perhaps more the result of careful study 

 of the features that constitute specific differences in the violets than 

 it is the segregation of already described species. At first glance the 

 peculiar violet described below would appear related to the Sa^^ittata 

 group, but a study of several sheets from various localities shows that 

 its affinities lie in an entirely different direction. 



Viola Novae-Angliae, sp. nov. Acaulescent. from slender, as- 

 cending rootstocks : earliest leaves ovate or triangular-ovate, cordate, 

 crenate, rounded or obtuse at the apex, 1-2 cm. long, on petio'es 

 2-4 cm. long; later leaf-blades thin in texture, triangular-lanceolate, 

 4-6 cm. long, 1.5-2.5 cm. broad at the base, tnpering to an acute 

 apex, the base subtruncate to cordate, margin crenate-dentate at the 

 base, obscurely and distantly crenate toward the apex ; dark green 

 and nearly glabrous above, paler and more or less pubescent beneath, 

 especially on the veins; petioles 7-15 cm. long, pubescent with white 

 tloccose hairs or glabrate in age: Howering scapes 6-10 cm. long, 

 not surpassing the leaves, bibracteolate below the middle, bractlets 

 ovate with subulate tips i mm. long : sepals oblong-lanceolate, 7-8 

 mm. long, glabrous, obtuse, 3-nerved, the basal auricle less than i 

 mm. long, rounded or truncate: petals narrowly oblong, 12-15 mm. 

 long, deep purple-blue, spreading, the lateral and lower ones densely 

 villous with white hairs, the lower one veined with dark purple: cleis- 

 togamous flowers short-sagittate and blunt, 3-4 mm. long on short 

 horizontal peduncles, 15-25 mm. long, their capsules subglobose, 

 about 6 mm. long, or less. 



Sandy shore, Fort Kent, Aroostook county, Maine, M. L. Fernald, 

 no. 2245, June 15, 1898. The type is sheet no. 338,658 in the 

 National Herbarium. 



The systematic position of this species is difficult to decide. By 

 its small cleistogamous flowers on horizontal peduncles it is related 

 to V. affinis LeConte. Its leaf-shape is somewhat suggestive of V. 

 emar^itiata (Nutt.) LeConte, but its dense pubescence removes it 

 from close relation with either of these species. 



In addition to the type mentioned above the following specimens 

 may be referred here. 



