2o-2 NATUliAL HISTORY OF PL.INTS. 



7. The transformation of certain stamens into sterile petaloid blades.' 

 — This transformation is not of generic value, for there is no genus 

 in which all the species present it. In Aberemoa, in Uiiona, it occurs 

 in a single species ; as well as in a few American species of Xylopia. 

 It takes place, as we have seen, sometimes with the outer stamens, 

 sometimes, but far more rarely, with the inner ones." But in this 

 order it does not seem to be due to cultivation. 



S. The conformation of the upper part of the receptacle. — We 

 may distinguish the form of complete concavity, with epigynous 

 insertion of all the Horal appendages exterior to the pistil, from that 

 in which the concavity is restricted to the summit of the receptacle, 

 or to a region not involving the insertion of the perianth, which in 

 this case is always hypog^'nous and inserted below the androceum. 

 Accordingly, while complete concavity has sufficed to mark out one 

 particular series, the Eupomatiea, the partial deformity may vary 

 from species to species in one and the same genus. The deep sac 

 on whose outer surface the stamens are inserted in most species 

 of X//Iojna may become a slight pit, or even a j)lane surface in 

 some.^ The same diversity is found in Artabotri/s* Hexalohus, &c., 

 though in these genera the cavity is never so marked as in certain 

 Xyhjjias. 



9. The ascending or descending direction of the ovules. — It will 

 be seen that this has no more importance here than in anj'' other 

 group, when we have to deal with numerous ovules arranged along the 

 whole length of the ventral angle of the ovary. In the same species, 

 in the same ovary, here as elsewhere we find ovules nearly horizontal 

 at the centre of the placenta, while they are more or less oblique, 

 ascending or descending, as they approach the top or bottom of the 

 cell. But when the ovules are solitary or few in number we do not 



' See Adfinsonia, viii. 32G. (p. 219) ; and X. malayana HoOK. F. & TiiOMS., 



- This peculiarity has been observed in the in this respect aflbrds a transition between these 



genus Anaxagorea alone (p. 207). and the other species of Xylopia, its receptacle 



^ Trof. Oi.iviu, in his enumeration of the being like an elongated cone solid for about 



Anonacfcc\\\i\\cFI(imofTro2)icalAfrica{^\.^0), two-thirds of its height, with the upper third 



an unpublished work, of whicrh he has kindly fa- alone hollowed into a shallow pit to receive the 



voured me witli the proof sheets [this work insertion of the carpels. 



was jiublishcd in 1808], has had no hesitation •* Especially in the species of tbc sectionParrt;-- 



in referring Mtlodorum africanum Kekth., to tahotrys, such as P. hexagyna MlQ. The sur- 



Xylopia, despite its convex receptacle. This face on which the carpels are inserted is flat, 



bna nearly the same form in most species of but it rises in a circle all round, projecting to a 



Jlabzelia, which wc have included in Xylcpia fair height in proportion. 



