538 PHANEROGAMIA. 



manifest underneath, and confluent into an intramarginal vein : they 

 vary from 5 or 6 to 15 lines in length, and from 3 or 4 to 12 in 

 width. Peduncles equalling or often exceeding the leaves when one- 

 ftoiuered, usually more elongated when three- (rarely fi ve-)floicered ; in 

 the latter case the main peduncle, an inch or so in length and slender, 

 is flattened (as in the other several-flowered species) ; the intermediate 

 flower sessile or nearly so; the others on pedicels of 3 to 7 lines in 

 length. Bractlets under the flower minute and very caducous. Flowers 

 nearly as large as those of the Common Myrtle. Calyx almost gla- 

 brous; the tube obovoid-turbinate ; the lobes 4, rounded, imbricated 

 in aestivation, mostly a little ciliolate, as are the four rounded-obovate, 

 deciduous petals. Stamens very numerous : filaments filiform : 

 anthers oval. Style filiform : stigma simple. Ovary usually three- 

 celled, sometimes two-celled, with the placentae in the axis. Ovules 

 numerous in two ranks in each cell, hemitropous, slightly reniform. 

 Berry globose, pulpy, 5 or 6 lines in diameter, apparently purple, 

 crowned with the persistent lobes of the calyx, ripening from 3 or 4 

 to 9 or even 12 seeds. Seeds lh to 2 lines in diameter, compressed, 

 orbiculate, often excised at the hilum ; the testa membranaceous, 

 chestnut-coloured, smooth, not adnate to the embryo. Radicle very- 

 long, cylindrical or somewhat club-shaped, incurved. Cotyledons 

 broadly oval, accumbent to the semiannular radicle, fleshy but rather 

 thin and flat, plane or sometimes slightly curved laterally, not cohe- 

 ring with each other or at all conferruminate, shorter than the radicle. 

 In all the seeds examined of var. /?. the radicle is truncate, as if the 

 extremity were cut off nearly in a line with the summit of the 

 cotyledons, instead of continuing so as to half surround them or more. 

 Taken with the difference in foliage, this would indicate the Eugenia 

 apiculata, Hook. & Am. as a distinct species ; but I suspect that the 

 peculiarity is a casual one. 



The plants here brought together pretty clearly belong (with the 

 possible exception indicated above) to one somewhat variable species, 

 which is doubtless the Myrtus Cheken of Feuillee, who first made 

 the plant known. Probably it likewise includes the Myrtus Luma 

 of Molina, to which Dr. Hooker's Eugenia apiculata var. Arnyan 

 [Arrayuhf] would very well correspond, having broadly oval leaves 

 and one-flowered peduncles. The only question is as to its size, and 

 the nature of the wood : the Luma according to Molina being a tree 



