1917] Fernald,— Gentiana clausa 147 
—— August 11, 1916, Fernald & Long, 14,068. New Hamp- 
RE: Androscoggin River, Shelburne, aie IL, 1882, Walter Deane; 
roadside west of Gates Cottage, Shelburne, September 12, 1907, 
A. S. Pease, no. 10,798; climbing high over trees by the Androscoggin, 
Shelburne, September 27, 1916, A. S. Pease, no. 16, r Mas 
comp Lake, Enfield, August 22, ‘1878, H. G. Jesup. shectnumien: 
Georgetown, August 9, 1907, E. F. Williams; Sudbury, September 2, 
» C. W. Swan; roadside, North Wilbraham, May 16 & 18, 1913, 
M. L. Fernald & F. W. Hunnewell, 2nd; banks of Connecticut River, 
Hadley, July 1, 1874, H. G. Jesup. Connecticut: banks of Con- 
necticut River, South Windsor, October 8, 1890, E. Watson 
Although resembling V. Labrusca in its usually continuous tendrils 
and inflorescences, in the contour of the leaf as well as in the rufescent 
tomentum of the very young leaves, and in its large fruits and seeds, 
V. novae-angliae in no material examined shows any tendency to retain 
the tomentum as does V. Labrusca, except as an insignificant vestige 
along the nerves on the lower side of the leaves. Its fruit has a clear, 
sharp acid flavor and quite lacks any suggestion of the “ muskiness” 
so characteristic of V. Labrusca. If the species were a hybrid of V. 
vulpina and V. Labrusca it is almost inconceivable that this peculiar 
flavor so characteristic of V. Labrusca should not appear in V. novae- 
angliae. 
V. vulpina, which V. novae-angliae resembles in its green foliage 
-and in its habitat in rich river-alluvium, has the more elongate leaves 
jagged-dentate with prolonged teeth; the young growth not rufescent; 
the tendrils and inflorescences with much more interrupted distribu- 
tion; and the berries and seeds decidedly smaller. 
VI. GENTIANA CLAUSA A VALID SPECIES. 
In the Synoptical Flora of North America Gray clearly defined the 
corolla-characters separating Gentiana Saponaria L. and G. Andrewsii 
Griseb. In the former species, as stated by Gray, the corolla is “light 
blue, an inch or more long, its broad and roundish short lobes erect, 
little and often not at all longer than the 2-cleft and many-toothed 
intervening appendages”; while in G. Andrewsii the corolla is “as 
the preceding but more ching and the lobes obliterated or obsolete, 
the truncate and usually almost closed border mainly consisting of the 
prominent fimbriate-dentate intervening appendages.” ! 
1 Gray, Syn. FI. ii. pt. 1, 122 (1878). 
