MYRTACEiE. 



545 



which has short and flattened filaments, bearing adnate anthers, a 

 completely bi-trilocnlar ovary, and with the cells becoming more or 

 less perfectly bilocellate by the growth of a false partition from the 

 back of each cell to meet the prominent placentae, much as in Rhodo- 

 ■myrtus (the character of which is somewhat weakened by this) ; — in 

 fact differing from Eumyrtus just as true Vaccinium and the section 

 Vitis-Idcea differ from the section Cyanococcus, Gray, Chlor. Bor.-Am. 

 p. 53. Turczaninow (loc. supra cit.), remarking these characters 

 (except that he does not notice the dorsal septum that meets each 

 projecting placenta), has founded on the flowers of this plant his genus 

 JJgni. But his character of the embryo must have been derived from 

 some other plant, or else it is incorrect : for the seed and hippocre- 

 pical embryo in my specimens of Myrtus JJgni, from C. Gay's Chilian 

 collection, are just as in Myrtus nummularia, &c. ; the testa is almost 

 bony and reniform, the radicle is long and slender, and the cotyledons 

 are short, narrow, and semicylindrical, not at all conferruminate. 

 Turczaninow may have had the seed of some species of Luma (vide 

 p. 535) : yet in these the radicle is not short, but remarkably long, 

 nor are the cotyledons truly conferruminate, nor the testa hard and 

 crustaceous, in any species that I have examined. 



♦ 



J Species dubia. 



5. Myrtus ? tenuifolia, Smith. 



Myrtus tenuifolia, Smith, in Trans. Lion. Soc. 3, p. 280 ; DC. Prodr. 3, p. 241. 



Hab. Near Sydney, New South Wales. 



This has scarcely been noticed, so far as I am aware, since the time 



Feuillee, the figure being doubtless reduced in size. For the leaves are represented of 

 barely 3 lines iu length, while in the description they are said to resemble those of the 

 Myrtle of Tarentum, and to be 7 or 8 lines long. A true congener, undoubtedly, is the 



Myrtus Candollii, Barneoud, in Gay, Fl. Chil. 2, p. 382, from the province of 

 Chilbe ; which I possess with undeveloped flowers only. It is pentamerous, and has the 

 same short, oblong or sagittate and apiculate anthers, and short filaments as M. Ifgni, the 

 filaments not exceeding the anthers in length in the full-grown flower-buds j but they 

 are not dilated nor flattened. It therefore further invalidates the character on which 

 Turczaninow most relies to distinguish his genus Uyni. 



137 



