MYRTACE^. 559 



Var. y. Vitiensis; ramulis foliisque ellipticis oblongisve glabris; in- 

 florescentia calycibusque plus minus cano-sericeis ; flwibus nunc sub- 

 pedicellatis. (Tab. 68.) 



Hab. Var. a. and /3. Tahiti and Eimeo, Society Islands ; common 

 on ridges, y. Feejee Islands; Ovolau, Muthuata, and Sandalwood 

 Bay; at the elevation of from 100 to 2,000 feet. — A Metrosideros, 

 mentioned by Dr. Pickering, as perhaps that of the Feejees but not 

 that of Tahiti, was noticed on the mountains of Tutuila, one of the 

 Samoan Islands, at an elevation of 2,000 feet; where it forms "a 

 spreading tree, 30 feet high, with the trunk a foot in diameter; the 

 leaves smooth ; the flowers scarlet, but not very showy." There are 

 no specimens of it in the collection. 



The M. villosa of Smith, with the young leaves, brancldets, and espe- 

 cially the inflorescence canescent with a villous-tomentose down, passes 

 so completely into glabrate and truly glabrous forms that it becomes 

 advisable to restore the earliest specific name, that of the elder 

 Forster. One of the intermediate varieties is doubtless the M. diffusa 

 of Hooker and Arnott, not of Smith ; the latter being a New Zealand 

 species. The leaves vary from ovate or oval to elliptical, or even ovate- 

 lanceolate, either rounded at both ends or acutish, or the base occa- 

 sionally subcordate; the copious pinnated veins manifest on both 

 sides but slender, minutely reticulated. Petioles lb to nearly 3 lines 

 long. Cymes sometimes axillary towards the summit of the branches, 

 usually terminal and in pairs, as described by Smith, but often soli- 

 tary or else three together at the summit. Flowers sessile or nearly 

 so, in threes or fives, at the summit of the partial peduncles ; which 

 distinguishes the plant from the Sandwich Island species. Stamens 

 red. Ovary three-celled, enclosed in the tube of the calyx, to which 

 its lower half is adnate. The mature capsule projects one-half 

 beyond the crateriform calyx (the lobes of which are tardily deci- 

 duous), to which barely its base remains adherent, and not very 

 firmly. 



From the above I am unable specifically to distinguish the speci- 

 mens from the Feejee Islands; var. y. They mostly have rather 

 narrower, elliptical or oblong, or even lanceolate-elliptical leaves, more 

 or less narrowed at both ends, in some specimens, however, as broad 

 as in the ordinary Tahitian forms, rarely inclining to obovate, all of 



