TEREBINTHA CE^. 



269 



Tapirira (fig. 302, 303) is also closely allied to Sorindeia by the 

 diplostemouous flowers aud unicarpellary gyufeceum, with, the same 

 insertion of the descendent ovule; but they have an imbricated 



Tapirira [Odina] Wodier. 



Fig. 302 Male flower {\). 



Fig. 3C3. Longitudinal section of male flower. 



corolla like Schiiius ; and, on the other side, if we only consider the 

 gynseceum of the male flowers, which, although sterile, is formed of 

 fom- or flve carpels in great part independent, they afi'ect the closest 

 afiinities with the Spondias^ especially those of the sub-genus Fou- 

 pariia (fig. 258), only differiag from them by a unilocular ovary and 



Gluta Benghas. 



Fig. 30-1. Hermaphrodite Fig. 307. Hermaphrodite flo-sver, Fig. 306. Longitudinal section of 

 flower (3). the perianth removed. hermaphrodite flower. 



fruit. They are trees and shrubs, sometimes climbing, with impari- 

 pinnate leaves, having 4- or 5-merous flowers, which have been ob- 

 served in all tropical regions of the globe, except Australia.^ 



Gluta (fig. 304-307) has very nearly the same organisation 



' We can only doubtfully place near Tojn- 

 rira, Sctmatoslaphis liarteri, a small tree from 

 western tropical Africa, with compound impari- 

 pinnate leaves, whose male flowers arc trime- 



rous, with imbricate diplostcmonous perianth, 

 iind the fruit a drupe with hard stone, containing 

 a descendent exalbuminous seed. The female 

 flower is unknown. 



