164 APPENDIX. 



sessilis and pedunculata with 4 petals either be- 

 long here or to Elwertia, unless with other ano- 

 malies. Jussieu deems the caps, uniloc. in all. 



45. COFFEA L. &c, this G, now greatly in- 

 ereased,includes at least 2 others blended G. Po- 

 tima Pers. with monosperm berry and the next. 



46. HEXEPTA Raf. (6 or 7) diff, Coffea, cal. 

 6-7dent. cor. 6-7fida, stam. 6-7, baccis sepe an- 

 gulatis 2sp. vix arillatis- -types 2 shrubs of East 

 Africa. 



47. Hexepta axillaris Raf. Coffea zangue- 

 barica Lour. M. <$c, fol. ovatolanc. ped. axil, 

 unifl. fascicul. baccis obi. ang. nervosis. 



48. Hexepta racemosa R. Coffea do Lour. 

 M. &c fol. ovatolanc. tuberc. pedunc. 4gonis ra- 

 cemosis. 



49. PERSIMON Raf. add to Mabola 21, the 

 Diospyros virginiana is stated to have 16 sta- 

 mens in two rows, while real Diospyros lotus 

 &e, only 8 in one row, if so which I will soon 

 verify, it must with other American sp. form the 

 G. or subg, Persimon, a very good name nearly 

 Greek in euphony although American. 



50. Add to 837, there is a previous G. Evos- 

 ma, Shrub of Australia and Lysianthides ; there- 

 fore the Evosmtis of Nuttal must be changed, I 

 propose EVELYNA, dedicated to Evelyn the au- 

 thor of a Sylva. 



51. Add to Pleuteron 673, some of the Brey- 

 nias with 6 stamens and double calix, were cal- 

 led Hermupoa by Loefling, the type had scar- 

 let flowers, compare my New Genera. 



52. TETRACERA G. in utter confusion by the 

 medley of G. thrown into it, Delima style 1, Pi- 

 ripea dioical &c, Euryandra 3 styles, Dolio- 

 carpus, Mappia, Calinea, Valbomia &c, which 



