212 Rhodora [DECEMBER 
Hook, Fl. Bor.-Am. ii. 140 (1838); Torr. Fl. N. Y. ii. 178 (1843); 
Gray, Man. 409 (1848); Fernald, Ruopora, iv. 150 (1902). E. 
purpureum Raf. New Fl. pt. iii. 50 (1836) in part.— Arctic and boreal 
regions, extending southward in peaty soils to Newfoundland, Nova 
Scotia, the Maine coast, the mountains of northern New England and 
northern New York; Pictured Rocks, northern Michigan (G. H. Hicks); 
Pipestone Valley and Lake Louise, Alberta (Stewardson Brown) ; Selkirk 
Mts., British Columbia (J. M. Macoun); Mt. Rainier, Washington 
(E. C. Smith, C. V. Piper) and Crescent City, California (Howell). 
Extremely variable in the length and breadth of leaves, which range 
from linear to elliptic in outline and from 2.5-7 mm. in length. The 
branchlets, too, are sometimes nearly glabrous, with only minute 
glandular puberulence, but in our northern regions and the Northwest 
quite as often minutely pilose with sordid or viscid, not white, hairs. 
The seeds of the more northern material are commonly about 2 mm. 
long, ranging between 1.8 and 2.6 mm. and in very rare cases to 3 mm. 
in length, while toward the southern edge of the range the seeds are 
frequently smaller, from 1.4-1.8 mm. long. In much of the small- 
seeded material the leaves run decidedly shorter than in most of the 
more northern plants and upon first studying the group we inclined 
to separate as a southern variety the plants with shortest leaves and 
smallest seeds. The study of a fuller series of specimens shows, how- 
ever, that no satisfactory line can be drawn either upon the basis of. 
length or shape of leaf or size of seed. The most extreme illustration 
of this lack of concomitance in these characters is a sheet of specimens 
from the Mealy Mountains, Labrador (coll. Dr. 4. P. Brown), with 
leaves only 3-4 mm. long, but with a fully developed seed 3 mm. long, 
the largest seed measured by us in the species. 
la. Var. PURPUREUM (Raf.) DC. Prodr. xvi. pt. 1, 26 (1869); 
Simmons, Vasc. Pl. Fl. Ellesmereland, 43 (1906); Robinson & Fernald 
in Gray, Man. ed. 7, 551 (1908) in part. E. purpureum Raf. New Fl. 
pt. iii. 50 (1836) in part. E. rubrum, Durand, Proc. Acad. Sci. Phila., 
1863, 95, not Vahl.— Northwestern Greenland, Ellesmereland, and 
Labrador.— A very little-known plant, resting upon an insecure basis. 
Rafinesque, assuming a larger knowledge of the northern flora than 
was at all justified, described his complex E. purpureum as follows: 
“EMPETRUM PURPUREUM Raf. E. nigrum Mx. and all our Amer. 
botanists, not of Lin. and European bot. E. rubrum Lapilaye fl.— 
Procumbent smooth, leaves scattered crowded, lower patent, upper 
imbricate, oblong-linear sessile uninerve obtuse flat on both sides, 
thickish, berries purple, sessile equal to the leaves and costate.— in 
Canada, Labrador, Newfoundland, White Mountains, Lake Superior, 
near the rocky shores. Michaux who first noticed this blended it 
