156 MR. G. BEKTHAM ON MYRTACE^. 



to what is believed to be that of Osbornia. The seed and embryo 

 are nearly those of some Australian Myrtiy the testa thin, the 

 radicle very long and much coiled round the linear cotyledons. 



Decaspeemum, Forst., is the name which ougbt to have been 

 adopted for the small genus which stands in our ' Genera Planta- 

 rum' as JV^elitris, Glaertn. In this we had followed Lindley, De 

 Candolle, and other modern botanists, not suspecting that their 

 identification of Gsertner's genus was erroneous. We had, un- 

 fortunately, overlooked the fact that Thwaites (Enum. PL Zeyl. 

 153) had ascertained that Gaertner's plant belongs to the Eu- 

 biaceous genus now known under the name of Timonius, Eumph. 

 A careful study of Goertner's description of the fruit w^ould, in- 

 deed, have shown that it could not be that of Decaspermiim ; but 

 what probably originally led to the error is, that he expressly de- 

 scribes the seed as exalbumiiious, and it was not until A. Gray's 

 careful review of the Guettardese that it was observed that the 

 albumen was wanting in Timo7iitis, or, indeed, that it was supposed 

 that it ever was deficient in any Eubiacea. For the Myrtaceous 

 genus we must therefore have recourse to Forster's name, al- 

 though far from appropriate. It consists of very few species, from 

 eastern tropical Asia and tropical Australia, nearly allied to 

 Myrtus. They have the hard testa and the embryo of the section 

 Eumyrtus ; but the inflorescence is racemose, as in some Ame- 

 rican Eugenics^ never cymose as in the many-flowered Myrti; the 

 ovary is 4- or 5-celled, whilst it is only 2- or 3-celled in Myrtm ; 

 and the ovules, 2 or 3 in each cell, or very rarely more, show an 

 approach rather to 2Iyrcia than to Myrtiis. 



Mybcia, DC, is a very large tropical and subtropical Ame- ^ 

 rlcan genus, allied to Myrtus^ but originally separated on account 

 of the embryo, which, in the few seeds then known, showed large 

 broad cotyledons, more or less folded. IN'ow that many more 

 seeds have been examined this character proves to be not near 

 so constant as had been supposed ; the embryo often varies from 

 species to species, and in some of the section Luma of Myrtxis 

 it is very much like that of some species of Myrcia ; and it is pro- 

 bable that if the seeds of all were know^n the connexion would 

 be found still closer. The genus may, however, be retained ; for> 

 besides the inflorescence, which is usually more compound, there 

 appears to be a more constant difference, in the ovary containing 

 onlv 2 collateral ovules in each cell. The cells are usually 2, or 

 rarely 3, and the flowers almost always 5-merous, as in the section 

 Eumyrtus. The 500 supposed species of Myrcia may probably 



