1896 SALIX CORDATA X SERICEA 393 
show more or less silkiness of capsules, and where there is silk- 
iness of capsule or leaf, even but sparingly, the problem is easy. 
The evidence now at hand, however, compels the extension of 
the idea of contamination. By taking account of every possible 
character, we find that hybrids, or rather intermediate races, 
include many examples having capsules perfectly smooth, and 
all but the youngest leaves likewise perfectly smooth. 
Upon the S. cordata side of our problem it is perhaps impos- 
sible as yet to form a proper conception as to its typical char- 
acter, either from description or observation. An intelligible 
and comprehensive description of it is yet to be written, and 
must be founded upon specimens collected from regions where 
S. sericea is not found. 
Fortunately in the characters of S. sericea we have rigid 
stability, a fact which furnishes us an invaluable guide to a 
partial exploration of the mazes of hybridization. Having exam- 
ined specimens from Massachusetts, New York (Ithaca), southern 
Pennsylvania, Ohio, Michigan ( Detroit and Port Huron), Illinois 
(Chicago), Missouri (Pilot Knob and St. Louis) ,and the Rocky 
mountains, I have found no noteworthy variation. Specimens 
of S. cordata were examined from New Brunswick, Niagara Falls, 
Ohio, Illinois, Iowa, Missouri, Minnesota, Colorado, Idaho, 
Oregon, California, and Kew Gardens. 
Notwithstanding the variability of S. cordata, taking a general 
view of it together with its varieties, including var. vestita and 
its hybrids with S. sericea, there is presented as a whole such an 
ensemble of stem, spread, color, foliage, flower and fruit as to 
enable one to distinguish it easily from all other species. The 
interest in this paper must lie therefore rather in biology than 
iM systematic botany. : 
: Assuming but one form of S. cordata, and one good variety, 
Viz, var. vestita, and knowing the freedom with which each 
Crosses with S. sericea, we shall have, theoretically, the following 
Combinations: cordata x sericea, cordata X vestita, sericea x 
weatita, and cordata < sericea X vestita. I think these combina- 
tions are recognizable in the vicinity of St. Louis. But a more 
