WNNEAN SOCIETY OF LONDON. XXSVU 



the extremities bearing a fertile anther fixed extrorsely upon the 

 tip of the segments — a very important feature. The ovary is quite 

 inferior, crowned on its outer edge with five thick triangular 

 sepals, which are valvate in aestivation ; it is from 5- to 12-celled, 

 each cell containing two or four superposed collateral ovules fixed 

 in the axis. The indehiscent fruit is a depressed globe umbili- 

 cated in the centre, where it is crowned by the persistent sepals ; 

 it has a more or less thin coriaceous pericarp divided by distinct 

 dissepiments into cells varying in number in the several species ; 

 in most cases only a single seed is perfected in each cell, which is 

 oblong, compressed, and reniform on one margin where it is at- 

 tached to the axis of the fruit ; and upon its reniform sinus a 

 broad cicatrix is seen, denoting the place of its adhesion to the 

 angle of the dissepiments — a feature hitherto unnoticed ; the seed 

 is covered by a very thin dark integument, which encloses an 

 exalbumiuous embryo consisting of two large fleshy cotyledons 

 and a short radical embedded within them at the ventral sinus- 

 All the plants of Napoleona are reduced to two species by the 

 authors of the ' Genera Plantarum,' and to one only by Profes- 

 sor Lawson; but in the present memoir many differences are 

 pointed out, in the habit of the plants, in the form and character 

 of the leaves, the colour and size of the flowers, the number of 

 parts in their whorls, the thickness of the pericarp in the fruits, 

 the number of cells, the shape of the seeds, the presence of pulp 

 (said to exist) in many, and its total absence in others — which 

 constant differences point to the existence of seven good species, 

 here described in detail. 



Upon the evidence thus brought together concevning Napoleona, 

 the author remarks that there is nothing in its structure to show 

 the slightest relation to Myrtaceae, that it is equally irreconci- 

 lable with the Barringtonieaa and with Lecythidese ; and in conse- 

 quence of these negative results we must search elsewhere for its 

 true affinity. This led the author to examine OmpJialocarpum, a 

 genus from the same region as Napoleona, and whose flowers 

 and fruit, of similar form, grow upon the trunks of the trees. 

 This genus has been generally regarded as belonging to Sapotaceae ; 

 but the authors of the * Grenera Plantarum ' place it in Teru- 

 stroemiacese. A full analysis of its flowers, and also of its fruit 

 and seeds, is here shown in detailed drawings, which seem to 

 prove beyond question that the genus belongs to Sapotaceae. On 

 comparing this structure with that of Napoleona, many unex- 



