PLANTS OF SEYCHELLES AND ALDAERA 287 



.longa. Embryo centralis, rectus, albumen fere sequans ; cotyledon es 

 foliacea?, cordatae ; radicula brevis. JJapaca Grijfitliii Hemsl. MSS. 

 in herbariis nonnuUis. 



Mahe : T. Risely Griffith, 1893 ; Thomasset, 157, 187. 



Apparently this tree is confined to Mahe, where, according to 

 Thomasset, it was formerly common in the mountains, though it 

 escaped Home and other early collectors. Specimens were first aent 

 to Kew in 1893 by the Hon. T. Risely Griffith, then Government 

 Administrator. 



These specimens were in fruit and my colleague Mr. N. E. Brown 

 provisionally referred them to JJapaca of Baillon, a genus restricted 

 to tropical Africa and Madagascar, In 1903 1 had to deal with 

 fruiting specimens sent by Mr. Thomasset and I provisionally named 

 the plant JJapaca Griffithii, and this name has a limited circulation, 

 though a description of the plant has not hitherto been published. 



In 1905 Mr. Thomasset sent to Kew excellent flowering specimens, 

 both male and female, and from their sti-ucture I was led to the belief 

 .that this tree could not be included in the genus JJapaca. A more 

 complete examination of the material has confirmed me in my opinion 

 and I now name it in memory of the discoverer. Apart from the 

 question whether the floral envelope is of the nature of an involucre 

 or of a perianth, there are differences which I consider of generic 

 value. But I regard the floral envelope of Riseleya as a true perianth. 

 On the other hand, there is no doubt that it is an involucre in JJapaca, 

 with several male flowers within each involucre ; each male flower 

 being provided with a small perianth and a central pistillode. The 

 involucre of JJapaca consists of five, or more, imbricate segments : 

 whereas the perianth of Riseleya both of male and female flowers is 

 composed of four parts in opposite valvate pairs. The stamens are 

 grouped in almost linear, not circular, clusters, opposite the perianth 

 segments, with a very small (nectariferous ?) scale between the stamens 

 and the segments. The centre of the concave hairy torus is otherwise 

 bare. 



I have placed Riseleya near JJapaca more, perhaps, on account of 

 its former association than its real affinity, yet I do not know of any 

 better position for it among the genera having geminate ovules. 



Phyllanthus Schimperianus Hemsl. (UiipJtorhiacece). Arbor 

 communis (fide Thomasset) ramis rigidis rectis glabrescentibus ; 

 ramuli laterales floriferi, graciles, pubescentes, internodiis quam folia 

 multo brevioribus. Folia breviter graciliterque petiolata, tenuissima, 

 papyracea, disticha, oblonga vel elliptico-oblonga, usque ad 2 poll, 

 longa, sed plerumque minora, utrinque rotundata, apice nunc minute 

 apiculata nunc obscure emarginata, subtus pallida, venis tenuis«imis 

 eleganter reticulatis. Flores utriusque sexus intermixti ; masculi 

 pentandri ; feminei staminibus imperfectis muniti. Capsula ignota. 



Mahe: Cajjucin at 1000 ft., Diipont; without localitv, Thomas- 

 set, 28. 



Specimens of this RhyUanthus formed part of a small collection of 

 Seychelles plants made by Mr. H. P. Thomasset for the late 

 AV. Schinq^er, botanist to the German Deep-Sea Expedition, who sent 



