AIB. G. BEXTHAM ON MYETACE^. 127 



In Eucalyptese the cymes are altered to umbels, either solitary 

 or several together on a common peduncle forming a panicle — a 

 constant character even in the few species allied to E. tessellaris^ 

 where the inflorescence is disguised bj the umbels being reduced 

 each to two or three flowers, and crowded several together on a 

 common peduncle, or in a few other species, as in E. cxiynia^ 

 where^ the flowers being sessile, the umbel is reduced to a head. 



In Metrosiderese the inflorescence, almost always compound, is 

 often variable or irregular. Usually cymose, the cyme sometimes 

 passes into the umbel or into the raceme, connecting occasionally 

 in one genus the several inflorescences so distinct in other cases. 



The variations in size, shape, colour, persistence, &c. of bracts 

 and bracteoles, usually constant in species, and often much affect- 

 ing the general aspect of the plant, have been sometimes made 

 use of for generic distinction. It has, however, appeared to me 

 that they can never be safely relied upon, except as serving 

 for a purely artificial distribution of species, as, for instance, in 

 Darwinia, 



The flowers are more or less unisexual in many IMyrtacca?, 

 especially in the tribe Leptospermea? ; but although this is more 

 frequently the case in some genera than in others, the circum- 

 stance can never be taken as a generic character. The uni- 

 sexuality is never, perhaps, absolute, and varies much in degree, 

 even in different varieties of the same species. 



I shall now proceed to consider the several tribes and genera 

 adopted in our ^ Genera Plantarum,' and offer a few remarks on 

 the limits we have been induced to assign to them. 



Tribe I. Ciiam/ELAttcie.^. 



When Chama^laucicoe were chiefly known by the genera Ca- 

 lytlirix and Verticordia, and when all the species ranged under 

 Bceckea were supposed to have a perfectly 2- or 3-celled ovary, 

 the differences between Chamselauciea* and other Myrtacea) ap- 

 peared very striking ; and Lindley (Veg. King. p. 721) raised 

 them to a distinct Order, " on account of their very peculiar 

 aspect, which resembles nothing among Myrtleblooms, except 

 some Bseckeas, their remarkable abortive stamens, their simple 

 ovary, which never indicates a trace of being formed by the 

 adhesion of more carpels than one, and their pappose calyx." 

 Since then, however, the careful examination of all the varied 

 forms assumed by species of Bcec'kea and Ba?ckea-likc plants, 



