Blake Revision of Polygala 3 



yet proposed to split the genus on this feature alone. Of the sec- 

 tions recognized by Chodat, here taken as subgenera in view of their 

 strong characters and for convenience of subordination of their 

 minor divisions, six 1 occur in the region under consideration. The 

 first four Phlebotaenia, Badiera, Hebecarpa, Hebeclada are 

 chracterized by a crestless, beakless keel; the fifth Chamaebaxas 

 by a beaked keel; and the sixth Orthopolygala by a fim- 

 briate-crested keel. Their characters can best be considered sepa- 

 rately. 



1. Phlebotaenia. Two species, confined to Cuba and Porto Rico, 

 shrubs or even trees up to fifty feet high (according to Wright's 

 label), with large showy purple flowers in axillary racemes and large 

 very coriaceous very densely and dichotomously reticulate-veined 

 leaves. The sepals are free and caducous, the wings caducous, and 

 the keel beakless and crestless; the capsule 1- or 2-seeded, large, 

 broadly winged on one side. The type species was originally pub- 

 lished as a distinct genus by Grisebach under the above name, but 

 its characters, when the whole range of variation in the group is 

 taken into consideration, do not support this view, and one of the 

 technical characters on which emphasis was laid by Grisebach 

 the introrse pores of the anthers is by no means a peculiarity of 

 this group. Careful dissection of species of Hebecarpa and Hebe- 

 clada shows exactly the same introrse-subapical opening that is 

 found in Phlebotaenia. In the Genera Plantarum (i. 136) of Bent- 

 ham and Hooker, Phlebotaenia is characterized as having " petala 

 lateralia e carina libera et basi dissita, superiora minora minuta 

 vel 0." In these features, however, Phlebotaenia agrees with other 

 Polygalas. The upper petals (" petala lateralia " of B. & H.) are 

 not joined to the keel, it is true, but are firmly adnate to the 

 staminal tube; the lateral (" petala superiora " of B. & H.) are pre- 

 sent and larger than in most Polygalas, although still much smaller 

 than the upper. The removal of Phlebotaenia from Polygala to 

 such a degree in the Genera seems to be based on a misconception. 



2. Badiera. I have recognized eleven species, all West Indian, 

 several of which have not been examined. They form a habitally 

 very distinct group of shrubs or small trees with oblong to elliptic 

 coriaceous slightly veined leaves and very short axillary racemes of 



1 Including Badiera, treated by Chodat as a subsection of his section Hebe- 

 carpa. 



