1918] Fernald,— Epilobiums Sect. Lysimachion 31 
by thousands or tens of thousands of years, it has certainly been a 
sufficient time for the Sable Island plant to have become thoroughly 
fixed in its characters, and even if, many thousands of years ago, it 
may have originated as a hybrid, it has upon Sable Island intensified 
its characters and become a thoroughly constant plant. 
The case of this plant is exactly comparable with that of E. densum, 
var. nesophilum, discussed in this paper, the peculiar variant of Æ. 
densum found upon Newfoundland and the Magdalen Islands, where 
no true E. densum is found, but in those areas suggesting that it might 
have originated in the long-distant past by the hybridization of E. 
densum of the South and E. palustre of the North. Whether these 
plants have had such an origin is entirely problematical and it may 
as confidently be argued that they are local developments, which by 
insular isolation have become fixed entities, and are really the result 
of natural selection. Whatever the origin of these plants may be, 
they are now absolutely definite and consistent, and the Sable Island 
plant is here proposed as 
E. MOLLE Torr., var. sabulonense, n. var., habitu foliisque ut apud 
formam typicam; caulibus foliisque dense cinereo-pilosis, pilis ad- 
pressis incurvatis; capsulis cinereo-pilosis valde glandulosis. 
Habit and foliage as in the typical form: stems and leaves densely 
cinereous-pilose with appressed incurved hairs: capsules cinereous- 
pilose, copiously glandular.— Nova Scoria: swampy edge of fresh 
water pond at Life Saving Station No. 3, Sable Island, September 9, 
1913, Harold St. John, no. 1282 (type in Gray Herb.); Dr. St. John 
has examined material collected at the same station in 1899 by Prof. 
John Macoun (no. 21,189). 
IHI. EPILOBIUM GLANDULOSUM AND E. ADENOCAULON. 
Epilobium glandulosum Lehm.' has long been a somewhat baffling 
species to interpret, chiefly because of the small amount of material 
in American herbaria. Haussknecht ? in his Monograph took up this 
species and recognized it from the Bering Sea region southward to Cali- 
fornia and New Mexico and southwestward to Japan, also from Labra- 
dor, Newfoundland and Quebec and the Carolina Mountains. Tre- 
lease * restricted the species to Alaska and northwestern Asia but said 
“Forms too near this also in British Columbia” and “ young specimens 
1 Lehm. Pug. ii. 14 (1830); Hook. Fl. Bor.-Am. i. 206 (1833). 
2 Haussk. Mon. Gatt. Epil. 273 (1884). 
3 Trelease, Mo. Bot. Gard. 2nd. Ann. Rep. 99 (1891). 
