JOURNAL 



OF THE 



ARNOLD ARBORETUM 



Volume III OCTOBER, 1921 Number 2 



NOTES ON AMERICAN WILLOWS. XII 



Camillo Schneider 



a. SYSTEMATIC ENUMERATION OF THE SECTIONS, SPECIES, 



VARIETIES AND FORMS OF AMERICAN WILLOWS 



In a note "Ueber die systematische Gliederung der Gattung Salix" in 

 Oest. Bot. Zeitschr. (1915) 273, I pointed out how difficult it is to elabor- 

 ate a systematic arrangement of this genus representing the true relation- 

 ship of the different sections. I do not agree with the arrangement made 

 by Andersson (1867 and 1868) nor with the new grouping proposed by 

 Von Seemen (1903) for reasons given in my note referred to above; and 

 having studied, too, the Willows of Eastern Asia (see in Sargent, PL Wil- 

 son, in. 40-179 [1916]) I am convinced that some of the characters upon 

 which Andersson and Von Seemen based their arrangements cannot be 

 regarded as of great taxonomic value. 



The real relationship between the different sections enumerated below 

 cannot be shown clearly in a linear sequence. The first four sections 



i 



include the species with more than three stamens and may be considered 

 as forming a rather natural group known as Pleiandrae or Pleonandrae, 

 but the sect. Nigrae does not seem to be too closely related to the others. 

 Sections v to xxiii comprise the Willows with two stamens or with 

 only one as it is the case with the last section Sitchenses. The occurrence 

 of only one stamen in most of the flowers is also to be observed in Salix 

 Uva-nrsi which has now been placed by me in Sect. Herbaceae. 



Besides the number of three or more stamens the first four groups have 

 in the pale deciduous flower-bracts of both the male and female flowers 

 another peculiar character in common. The same kind of bracts 

 found in Sect. V. Longifoliae, a group of otherwise very different species 

 w r hich occupies an isolated position among the American Willows. 



Sect, vi to xxiii are arranged as far as possible according to the 

 impression I got of their relationship by comparing all their taxonomic 

 characters, I am not yet in a position to propose a good key clearly show- 

 ing how to distinguish all those sections. As I have already explained 

 in note vm (Jour. Arnold Arb. i. 229), I thought it best not to unite 

 species of apparently no close affinity in the same group but to propose 



