— 
1918] Fernald,— Epilobiums Sect. Lysimachion 37 
The overwhelming weight of opinion among post-Linnean botanists 
has been that the name E. alpinum belongs properly to E. anagallidi- 
folium Lam. Haussknecht cites no less than 125 European authors 
who have so treated the plant, and, since Haussknecht’s own work, 
such critical students of the European flora as Britten & Rendle and 
Druce have clearly designated E. alpinum as signifying E. anagalli- 
difolium. In view of this great weight of authoritative usage and 
the difficulty of proving clearly that the name E. alpinum applies 
more definitely to any other plant, we may well retain it in its long 
established sense for E. anagallidifolium, thus subscribing to one of 
the general principles (Art. v.) of the International Rules of Botanical 
Nomenclature that “when the consequences of rules are doubtful, 
established custom becomes law.” 
Moore, in the same paper in which he would identify £. Hornemanni 
as the Linnean EF. alpinum, states that the seeds of E. Hornemanni and 
E. lactiflorum “do not differ in any respect” (p. 142), arguing that 
Haussknecht could not have seen the seeds correctly. Nevertheless, 
examination of seeds from every mature sheet in the Gray Herbarium 
of these species shows conclusively that when seen against strong 
illumination the margin of the seed of E. lactiflorum appears very 
regular and smooth while the profile of the seed of E. Hornemanni is 
distinctly pebbled or, as Haussknecht says, “ papillose.” Only in the 
very rarest instances, and then in regions where both species abound, 
has there been any question in placing the plants in two definite species, 
one with the seeds smooth, the other with pebbled seeds. Further- 
more, all the material from Scandinavia, Greenland, Arctic America, 
and the region southward to the Shickshock Mountains and the 
White Mountains with the smooth seed of E. lactiflorum has the petals 
distinctly milk-white, as described by Haussknecht, or at most with 
a tinge of light pink, and in all these plants the expanded flowers 
(in dried condition) measure 3-5 (very rarely -6) mm. long. On the 
other hand, the material of E. Hornemanni, the species with pebbled 
seeds, has the expanded flowers 5-10 mm. long, the petals (except 
in rare albinos) lilac to rose-purple. The writer is therefore con- 
strained to regard the two, as was done by Haussknecht, by Trelease, 
and by numerous later students, as clearly distinct species and not, 
as Moore argues, merely color forms with no other differences except 
the color of the petals. n 
Besides E. alpinum (E. anagallidifolium), which occurs southward to 
