160 MB. G. BENTHAM ON MYRTACE-TS. 



reduced to Eugenia, The species are most extensively distributed 

 over tropical and subtropical America and Asia, with a much 

 smaller number in tropical and subtropical Australia and Africa ; 

 and a very few of tbem are cultivated for their edible fruits or for 

 their flower-buds used as spices. 



The chief character connecting this vast group lies in the em- 

 bryo, which is thick and fleshy, sometimes apparently homogeneous, 

 but in most cases showing more or less distinctly two large thick 

 cotyledons and an exceedingly short radicle. In other respects 

 the embryo varies in shape, straight or somewhat curved; the two 

 cotyledons are equal and hemispherical, or unequal and irregularly 

 shaped, closely united or separable or quite distinct. The great 

 difficulty, however, has been to find accompanying characters in- 

 dependent of the embryo, which is so seldom procurable, either in 

 wild or cultivated specimens ; for the 2-celled ovary with several 

 ovules in each cell, the small stigma, and the stamens are quite those 

 of Myrtus. The flowers of JEugenia arc almost always 4-merous, 

 and this was at one time thought to be a safe character to rely upon; 

 but, as already observed, several South- American species, which 

 on this account were transferred from MyrUis to Eugenia^ had, 

 when their embryo became known, to be restored to the former 

 genus, of which, indeed, they have the habit. Avery few species, 

 moreover, of true Eiigenice are exceptionally 5-merous ; and m 

 some, especially Asiatic ones, owing to the petal-like nature of 

 the inner calyx-lobes, or to the abortion or consolidation of 

 some or of all the petals, it is not easy to say what is their real 

 number. Inflorescence, as pointed out by Grisebach, is an excellent 

 indication for the majority of the American species. There are no 

 secondary cymes in the Ei(genice oi that continent. When the flowers 

 are clustered or paniculate in the axils it is from the contraction or 

 development of leafless flowering branches ramified like the leafy 

 ones ; whilst in Myrtus and Myrcia the secondary cymose inflores- 

 cence, if not developed, is generally indicated by the opposite brac- 

 teoles under the calyx. But this collateral character fails entirely for 

 the Asiatic species, the great majority of which have a trichotomous 

 cymose inflorescence. Here, however, there is no practical dini- 

 culty , as there are no true Myrti known from tropical Asia. Australia 

 has, like South America, several species of both genera, and hke 

 tropical Asia has both the inflorescences of Eugenia \ but all the 

 Australian Eugenics are 4-merous, and the Myrti 5-merous, except 

 M.fragrantissima^ T. MueU., which has every appearance of bcmga 

 4-merous Myrtus^ but remains doubtful as the seed is unknown. 



