1901] Williams, — Tree Willows at Fort Kent, Maine 277 
TREE WILLOWS AT FORT KENT, MAINE. 
EMILE F. WILLIAMS. 
Tune St. John river, where it forms the northernmost boundary of 
Maine between St. Francis and Hamlin, has already been most pro- 
lific in botanical rarities, but that its resources are by no means 
exhausted was demonstrated very forcibly during the short botanical 
trip made there last August by Dr. B. L. Robinson Mr. M. L. Fer- 
nald and myself. Many most interesting plants were collected and 
these will be noticed in due time in this journal, but I wish in this 
instance to call attention to some remarkable willows which we found 
growing in a hillside bog at Fort Kent. 
Salix discolor assumed here the habit and proportions of a fair 
sized tree. The trunk of one specimen measured forty-three inches 
in circumference at two feet from the ground. S. dalsamifera, which 
I believe has always been considered a shrub, here attained a diam- 
eter of fourteen inches at two feet from the base. Like S. discolor 
it assumed a tree-like habit and both these species were represented 
by specimens not less than twenty to twenty-five feet high. Another 
willow proves to be of more than usual interest. I collected speci- 
mens from a tree measuring seventeen inches in circumference at 
two feet from the ground and not less than twenty-five feet high on 
July 22nd, 1900, with Mr. J. Franklin Collins of Brown University. 
We referred these last winter to S. pentandra of Europe and north- 
ern Asia, but rather doubtfully as the station where they were col- 
lected hardly seemed likely to harbor introduced species. We paid 
a visit to these trees on August roth, 1901, and collected more mate- 
rial which has been critically examined by Mr. Fernald, who pro- 
nounces it to be S. Zucida, Muhl., var. macrophylla, Andersson. 
Andersson described this variety in his monograph of Salix (DC., 
Prodromus, XVI, Part 2, 205) from a specimen of Lyall’s, collected 
in 1859 on the Frazer river ( British Columbia) and from a specimen 
of Bourgeau's from Rio River (presumably in the Saskatchewan 
country). Fortunately there is a good full specimen of Lyall’s in 
the Gray Herbarium and it matches exactly our material. 
The important characters separating it from the ordinary S. /ucida, 
in which the leaves when mature are quite glabrous, are the closely 
pubescent branchlets with an only occasional tendency to become 
CONO 
