154; MR. G, BENTHAM ON MYETACE^. 



a few Soutb-Ainerlcan, chiefly Brazilian species, which, like Temttj 

 have the flowers and thin testa of Luma^ but differ slightly in the 

 deciduous calyx-lobes. 



4. PseudocaryopTiyllus, Berg, containing several South-Ame- 

 rican species (including some of Kunth's Myrti), which, on ac- 

 count of their larger, more coriaceous leaves, and tetramerous 

 flowers, have been referred by recent authors to Eitgenia] some of 

 them, of which he had" not seen the fruit, are even still placed 

 there by Berg. They prove, however, to have the ovary aud 

 seeds o^ Myrtus^ but differ from all the above-mentioned sections 

 in their numerous flowers in trichotomous cymes. 



5. MyrciantheSj Berg, is made up of four species, of which 

 two have furnished the generic character — Jf. cisplatensis^ Berg 

 (^Eugenia, Camb.), and 3L apiculata^ Berg- These appear to me 

 to be truly referable to the section Luina, Berg characterizes 

 them chiefly by the embryo with thick plano-convex cotyledons, 

 an exserted radicle, and a well-developed plumula enclosed be- 

 tween the cotyledons — that is, nearly the embryo of Eugenia, ex- 

 cept that in that genus the radicle, when elongated, is turned iu 

 against or between the cotyledons; and the development of the 

 plumula is so anomalous in the order, that one would be unwil- 

 ling to admit it without repeated verification. The habit also is 

 entirely that of the several-flowered Myrtiy and very different from 

 that of Eugenia. Unfortunately our specimens of all the species 

 are in flower only, and we have no seed to examine ; but, judging 

 from the figure in tbe * Flora Brasiliensis,' t. 82, we should con- 

 jecture tliat the so-called cotyledons may possibly be a very thick 

 radicle folded on itself, such as we have found it in some of the 

 JBlepharocalyx group, and such as is described by Lindley,in Paxt. 

 ri. Gard. iii. 149, in Eugenia apiculataj DC, a true Luma^ and that 

 the supposed plumula consists merely of the smaU inflected coty- 

 ledons. Of the two remaining species, M, hninnea, Berg, of wliich 

 the fruit is unknown, is probably a Myrtus ; M. edulis, Berg, of 

 which the plumula is not mentioned, may be a Eugenia. 



Of the nine Australian species of Myrtus four have the 5-me- 

 rous flowers, hard testa, and small cotyledons of Eumyrtus ; but m 

 three of tliem the embryo is very much longer and spirally invo- 

 lute, as in the Campomanesia group ; two, also 5-merous, have the 

 small cotyledons oi' Eumyrtus , the spirally involute embryo of Cam- 

 pomanesia^ but with a testa approaching more nearly to that ot 

 Luma, being of a rather thin consistence and remarkably granu- 

 lar rugose, and one of them has the dissepiment incomplete as m 



