MR. O. liENTHAAI OX MyKTACP:-E. 139 



each other but by vague characters of degree in the breadth aud 

 consistence of the leaves, in the indumentum of the inflorescence, 

 and in the colour of the stamens. 



Melaleuca, Linn., which, after Eucalyptus, is the largest genus 

 of capsular Myrtacese, is also a very natural one, so much so, 

 indeed, that all attempts to divide it into good subgenera have 

 failed. It is also very well defined by its exserted stamens united in 

 bundles opposite to the petals and bearing versatile anthers, only 

 passing into Callktcmon by a very few intermediate forms ; and 

 two species only have been proposed as separate genera : one, Jf. 

 teretifoliay Endl., is the Gymnagathis of Schauer, without any cha- 

 racter at all, for the inflorescence on which it is supposed to have 

 been founded is that of many other species ; the other, M. angtis^ 

 tifolia, Gsertn., or Asteromyrtus, Schavi., has, with two other 

 species, the calyx-limb falling oflf after flowering in a circumsciss 

 ring — a character unaccompanied by any other or by any difference 

 in habit, and therefore not available further than for an artificial 

 section. The coherence of the fruiting calyces, which was also 

 relied upon, is not constant in either of tlie three species. A few 

 species of Melaleuca are exceptional in the subtribe Euleptosper- 

 mese by their opposite leaves, but cannot constitute even a distinct 

 section, as they belong to very different natural series. One 

 species, the old M. leucaclendron, Linn., the only one which from 

 Australia spreads itself over the Indian archipelago and the Ma- 

 layan peninsula, is, with this very wide geographical range, also 

 singularly polymorphous. It has been divided into more than a 

 dozen species ; and most botanists retain two, three, or four as dis- 

 tinct, the extreme forms being widely dissimilar ; but the charac- 

 ters, derived chiefly from the shape and size of the leaves, from 

 the dense or interrupted spikes, from the size and colour of the 

 flower, and from the indumentum, are so variously combined in 

 different specimens, the forms at other times pass so gradually 

 one into the other, or differ so much at different ages, or even on 

 different branches of the same tree, that I have completely failed 

 in the endeavour to sort the specimens into distinct races. 



The seeds, in the few species where they have been examined in 

 the ripe state, differ considerably in shape, in the presence or ab- 

 sence of wings, and in the shape of the cotyledons of their embryo ; 

 but these differences, as far as known, do not appear to be available 

 for the distinction of sectional groups. 



Lamarchea, Gaud., and CoKOTnAM>TS, Lindl.,the former mo- 

 notypic, the latter consisting of two species only, diflfer from Me- 



