MB. G. BENTHAM OX MXETACEiE. 153 



but the habit, the 1-flowered peduncles, the ovary, and the seeds 

 are so decidedly those of the typical European Myrtus, that we 

 have followed A. Gray in reuniting it as a section. 



2. Eumyrtits, or the genus Myrtus as limited by Berg. Besides 

 our own European Myrtle, this includes a considerable number of 

 extratropical and Andine South- American species ; they are all 

 5-merous, the peduncles are 1-flowered, or very rarely 3-flowered, 

 the anthers are versatile, and the seeds,* as far as known, have 

 always a hard testa, a long curved but not spirally rolled embryo, 

 with very small colytcdons. 



3. Leantria, a sectional name taken up by A. Gray from a 

 proposed generic name of Solander's quoted by Eorster, is tlie 

 group established by Berg as his genus Myrteola, It contains 

 about ten species from Andine and extratropical South America, 

 which have either the habit of Humyrtus or are smaller and more 

 prostrate, and have also its characters, except that the dissepiment 

 of the ovary is incomplete, not reaching the top of the cavity. 



4. Luma was proposed by A. Gray as a genus under that 

 name, and by Berg under that of Myrceeugenia^ for several 

 species, chiefly Andine or Chilian, which, like the Uynis, had, 

 whilst their seeds wxre unknown, been referred to Eugenia 

 on account of their 4-merous flowers. They have nearly the 

 habit of EumyrtuSy but have frequently three to seven flowers on 

 the peduncles ; and the seeds show more or less of an approxi- 

 mation to those oi Myrcia) the testa is thin- and the colytedons, 

 larger in proportion to the radicle than in Eumyrtus and TTgni^ 

 vary from one species to another, narrow or broad, long or short, 

 flat or more or less folded. It was, on observing this great 

 diversity from species to species, and the combination of the thin 

 testa and small cotyledons in Temu and Blepharocalyx of Berg, 

 that we were prevented from following A. Gray in adopting the 

 genus Luma^ which, as a section, rests solely on the texture of the 

 testa ; and that character even Is invalidated by at least one 



Australian species. 



The following South-American Myrtoid groups had not come 

 under A. Gray's observation: — 1. Temu, Berg, a genus proposed 

 for one or two Chilian species, with the flowers and thin testa of 

 Luma^ but with the small cotyledons of Eumyrtus, thus closely 

 connecting the two sections. 2. Anamomis, Griseb., establislied 

 on three West-Indian species, which appear to me, in every 

 respect, to come within the section Luma, with which it did not 

 occur to Grisebach to compare them. 3. Blepharocalyx ^ B 





