144 MR. O. UENXnAM ON MTRTACE.E. 



propriety, be raised to fifteen, or, perhaps still better, reduced to 

 six. 



AciCALYPTTJS, A. Gray, was established for a species from the 

 Fiji Islands, to which afterwards a second was added from the 

 same locality, both of very doubtful affinity, being only known 

 from specimens in flower and bud. A. Gray, from the appearance 

 of the ovary, suspected that the fruit was- capsular, and on that 

 supposition indicated the affinity to Eucalj/pttts in its 4-merous 

 flowers and circumsciss operculum. This operculum, however, is 

 in both species evidently formed of the calyx alone, with the 

 free petals inside as in Calyptranthes^ not of the corolla alone or 

 combuied with the calyx as in JEiicahjphis \ and we therefore, in 

 the ' Genera Plantarum,' acting still on the supposition that the 

 fruit was probably capsular, placed it at the commencement of 

 Metrosidereae instead of among Eucalyptese. Since then Seemann 

 has discovered, amongst his Eugenias from the same islands, what 

 he presumes to be a third species, in which he finds the fruit to 

 be baccate ; and he therefore reduces the whole to Calyptranthes. 

 In this he may be right; but at present it can but be the result 

 of pure conjecture, the seed being unknown or at least unexamined ; 

 the appearance of the ovary and the 4-merous flowers, as well as 

 the geographical station, are against the union. The habit and 

 the arrangement of the petals in the third species {CaJyptranthes 

 eugenioideSy Seem.) are also much more those of Eugenia, sect. 

 Sy^ygium^ than of Calyptranthes . Our specimen has no fruit, so 

 that we can determine nothing; but should the seed prove, as is 

 probable, to have the Eugenia embryo, then A. Gray's genus Aci- 

 calyptics will stand under that name or be reduced to Cleistocalyx ot 

 Blume, but must be transferred to Myrtese, next to Eugenia. 



Tristania, E. Br., with the 5-adelphous stamens of Melaleuca, 

 has the habit, inflorescence, and other characters of Metrosidereflp. 

 It differs from Metrosideros and its immediate allies, besides 

 the stamens, in the ovules tending downwards instead of up- 

 wards, and usually in its alternate leaves. It comprises, however, 

 three sections, having almost as strong claims to be considered 

 distinct genera as the small genera more closely allied to Metro- 

 sideros^ but which it appears more convenient to follow Brown m 

 retaining under one generally adopted generic name. The most 

 distinct is NeriopTiyllum^ in which the leaves are opposite, the 

 ovules are numerous and mostly horizontal, and the union of the 

 stamens in bundles is less decided than in the other sections. It 

 consists, however, but of a single Australian species, and no object 



