MR. G. BEXTHAM 0>' MYETACEiE. 159 



angle of the cells. Berg proposes to limit the genus to the single 

 P. communis^ Lindl. (P. officinalis^ Berg), ^vhich has 4-merou3 

 flowers and a spirally involute embryo of 2 or 3 coils, whilst the 

 other species, which he separates under the n^raeo? Amomisy have 

 5-merous flowers and a much less involute embryo; but as w^e 

 have not admitted these cliaracters as sufficient in Myrtus^ where 

 the large number of species show their want of conformity witli 

 habit and other characters, so can we much less agree to their se- 

 parating in this instance a single species without any difference in 

 habit, AVe have found the embryo of Pimenta acris {A^nomis 

 acris, Berg) intermediate between that of P. communis and the 

 other species of A^no mis of Berg, 



MvRRniNiUM, Schott, independently published also as Feliciana 

 by Cambessedes, and as Tetrastemon by Hooker, is a single species, 

 widely spread in South America, which at first sight appeared so 

 anomalous as to have been placed with some Melastomaceae in the 

 now abandoned order of Memecyleae. The stamens w"ere supposed 

 to be definite ; but, although very few in number, they vary from 

 four to eight ; they are not in a single series, and are not placed in 

 any regular position as to their alternation with sepals or petals, 

 thus showing all the characteristics of indefinite stamens. The long 

 straight filaments (which give them a peculiar aspect) are those 

 of ^anthosfemon, of Feijoay or of Eucalyptus cornuta and its allies. 

 Myrrhinium has also numerous ovules, placed on the margin of a 

 bilamellate placenta, which is rare in the tribe, but occurs in a few 

 species of Myrtus^ and is moreover an arrangement strictly ana- 

 logous to that of the ovules of BircTcea and XantJtostemon, which 

 are inserted in a ring round the margin of a dilated placenta. The 

 embryo of Myrrhinium is apparently homogeneous ; so that it is as 

 yet doubtful whether it consists of a radicle with minute or abortive 

 cotyledons as in some species which we include in Myrtus, or 

 whether the cotyledons are conferruminate as in Eugenia — a point 

 that can only be settled by watching its germination. 



Eugenia, Linn., is at once the largest and the widest-spread 

 genus of the Order, and the one which has occasioned the greatest 

 diversity of opinion as to its delimitation. Above 700 species are 

 described, which a careful scrutiny might reduce to about 500; 

 and whilst several eminent botanists, whose example we have 

 followed, retain the genus in its integrity, others of equal ability 

 have distributed the species into six or into ten genera, and others, 

 again, have endeavoured to establish nineteen, besides two or three 

 which we have adopted, but will perhaps ultimately be likewise 



