SALICIFLOR^:. 



337 



sidered as the youngest types, the strongly pronounced metamor- 

 phosis supporting this theory, as also the formation of the integu- 

 ment of the ovule, the one thick integument being undoubtedly 

 derived from the coalescence of two a holochlamydeous ovule, etc. 



The Apetalte and Choripetalae are united into one sub-class. The 

 leaves of the perianth in this case are, as a rule, free from each 

 other, the structure of the flowers presents many differences, and 

 the ovules have as a rule 2 integuments and a large nucellus. Con- 

 siderable uncertainty still prevails regarding the arrangement and 

 the relationship of the individual families of the Choripetalse, and 

 some of the following families are hardly quite natural ; but the 

 best arrangement arrived at so far has been adopted here. 



At the end of the book a review of the orders of the Dicoty- 

 ledons will be found. 



Sub-Class 1. Choripetalae. Petals free. 

 Family 1. Saliciflorae. 



Trees and shrubs, which, in the structure of the vegetative shoot 

 and the catkin-like inflorescences, resemble the Querciflora?, bat 

 the structure of the flower differs so much from them, that the 



FIG. 322. Male and female catkins of Sate caprea. 



only order brought under this heading Salicacece well deserves 

 to be separated and to form a family of its own, the nearest 

 relatives of which are still doubtful. As Juglandaceae and Myricacea? 

 also deserve to be placed in a special family, the name Amentacea (Catkin- 

 bearers), hitherto applied to all of these plants, cannot be retained as the name 

 of a family. 



There is only one order. 



w. B. z 



