199 
Of the inconstancy of S. cordata, Glatfelter (Trans. Acad. Sci. 
St. Louis, 7: 139, Jan. 4, 1896) states that of 184 specimens of 
supposed .S. cordata which he collected at random near St. Louis, : 
“13 were rejected as hybrid sericea and cordata either fully identi- 
fied or probable.” He also dwells upon the great variability of 
S. cordata and discredits Mr. Bebb’s species Salix Missouriensis. 
Professor Dudley spent much time studying willows about 
Ithaca. His conclusions (Cayuga Flora, p. 87), verified by Mr. 
Bebb, were that hybrids between S. cordata and related species 
were not infrequent. It has been a matter of some surprise that 
Professor Dudley or some other of the many students who have 
done field work about Ithaca did not discover these plants before. 
It hardly seems possible that they could have been passed over 
year after. It seems more probable that they have sprung up 
Within the last ten years, perhaps immediately after the attempt 
to reclaim the swamp was abandoned. 
Salix cordata as it grows about Ithaca is an exceedingly variable 
Plant. It presents several forms and grows along the creeks and 
in low grounds everywhere. 
The winter buds promise to aid much in the determination of 
species of willows Sa/ix cordata and S. candida have very different 
buds. In the former they are large and flattened upon the side 
next to the branch and appressed. They are acute, the apex fre- 
quently reaching the base of the next bud above. They are dull 
and frequently hairy. In the latter the buds are short, rounded at 
the apex and decidedly spréading. They are polished and shin- 
ing. The hybrids all have buds intermediate between the two 
Species ; they are more divergent, shorter, and more blunt than 
those of S. cordata, and vary from smooth to hairy. 
Not until the winter buds came into consideration did the ques- 
tion of the parentage of our hybrids seem satisfactorily solved. 
Plate 267 illustrates forms of buds of Sa/ix candida and S. cordata, 
and also the buds of the hybrids. The buds of the hybrid plants © 
have the gradually contracted upper portion of .S. cordata, but are 
inclined to be blunt. They are not always strictly intermediate 
between the two parents, but sometimes approach more nearly to 
one parent or the other. This variation is correlated with the re- 
lation of the hybrid to the parents as shown by other characters. 
It should be noted, too, in this connection that several spe- 
