— 47 — 



rSP, tfo:^5?3ff, I «TESTIS, (i^^^'IMf^ ^o (la H). 

 =*; = £6 ^ ffiM ****£■ ^ * -# ' P9S ' JK ? ^ /"«Si UK® = — 



'«f-Jo (® I). 



The structure of flowers, and the main characteristics for 



the classification of Salicacce. 



The development of the flowers of Salix. 



In Salicaceoe, two genera have been only enlisted in the books of 

 plant taxonomy. Griffith had distinguished Balsamiflua from Populus 

 by the shape of cupule, but no other botanists had ever regarded it 

 valid. From the genus Salix, Opiz separated Salix amygdaloides as a 

 distinct genus Gruenera, and Kerner founded another genus Chamistea 

 for the group of Salix reticulata. However, these two genera are not 

 accepted, and in consequence Populus and Salix are only the genera 

 remaining still in Salicacea?. They are usually classified as follows. 

 Buds with several imbricated scales. Perianth cup-like. ...Populus 

 ■ Buds with single scale. Flowers without perianth, but with 



nectary Salix 



Of these, Salix has more degenerated simple flowers. These simple 

 flowers are supposed to have been degenerated from tetramerous 

 flowers (Valenovsky in Beihefte Bot. Centralb. XVII, p. 123-128, 

 1904). In 1920, the writer obtained the female flowers of Salix 

 splendida, in which he found there neither perianth nor nectaries, but 

 deciduous stigmas which articulate with the styles. So, the writer 

 proposed to make this species as new genus Chosenia and distinguished 

 it from both Populus and Salix as indicated below. 



(Flowers with nectaries. Style one Salix 



iFlowers without nectaries 2 



Buds with many imbricated scales. Perianth cup-like. Style 



1 Populus 



Buds with one scale. Perianth none. Styles 2 distinct. Stigma 

 articulated with style Chosenia 



