Mr. J. Miers on the Affinities of the Icacinacese. 223 



are 2-lobed and 4-celled. This tribe will consist of the genera 

 Icacina, ApodijteSy Rhaphiolepis, Mappia, DesmostachySj Leretia, 

 and Poraqueiba. 



2. Sarcostigmea. These differ from the preceding tribe, in 

 the absence of a more or less elongated style : here the true 

 style, seen only in a young state of the ovarium, consists in a 

 broad, depressed, conical and more or less hollow process, having 

 a very small apical aperture, which is terminated by 4 or 5 ex- 

 tremely minute stigmatic teeth : with the growth of the ovarium, 

 this assumes a more fleshy, broader, and more discoid appearance, 

 and the stigmatic teeth become less visible, the whole soon taking 

 the form of a depressed, 4-5-lobed, sessile stigma. The ovarium 

 and fruit, in their structure and development, resemble what is 

 seen in the former tribe ; the stamens are also equal in number, 

 and alternate with the petals, and the anthers are likewise 2-lobe(i 

 and 4-celled. This tribe will consist of Pennantia, Stemonurus, 

 Sarcostigma, Discophora, and probably also Phlebocalymna, 



3. Emmotea. This at present is only represented by a single 

 genus, Emmotum of Desvaux, which is certainly identical with 

 Pogopetalum of Bentham, and which differs from all the others in 

 its plurilocular ovarium, and the singular structure of its anthers, 

 which are 2-lobed, and consist of 2 uniloc^llar, evalvate and boat- 

 shaped pollen-cells attached to a cordate and apiculate connective, 

 fixed extrorsely upon the reflexed point of the filaments in the 

 sinus of its anterior face, the pollen escaping by the splitting of 

 the dorsal margin of each valve from the posterior surface of the 

 connective, along the whole line of its attachment ; they vary 

 also in having an ovarium with three cells laterally placed in the 

 manner before mentioned. These peculiarities, so very opposite 

 in character to the features we invariably meet with in the other 

 tribes, very naturally suggest a doubt as to the propriety of re- 

 taining this genus in the order ; but no satisfactory conclusion 

 on this head can be entertained, until some information be ob- 

 tained respecting the structure of the fruit and seed. 



I take this opportunity of remarking, that I have lately exa- 

 mined with attention the features of several genera newly proposed 

 and described by Prof. Blume, in his ^Mus. Lugd. Bat.,^ and 

 referred by him to the Olacacec^. Among them is Nothapodytes 

 {loc. cit. p. 248), which will be seen to conform in all respects 

 with Mappia (Jacq.), a genus shortly to be described at length. 

 The characters given of his Pleuropetalum {loc. cit. p. 248), by 

 the same distinguished author, will easily be recognized as those 

 of Bursinopetalum of Dr. Wight, placed by that eminent botanist 

 in Olacace<2y but which I have shown must be referred to Aqui- 

 foliacecBy it being nearly allied to Villaresia : having to describe 

 in the sequel some new species of both these genera, I will then 



