134 MB. G. BEKTHAM ON MYETACE.E. 



other three sections. Of the two intermediate sections the anthers 

 of Harmogia approach nearest to those of Scliidiomyrhis^ and the 

 species are all eastern ; whilst the anthers of Oxymyrrhine tend 

 rather towards l^ahingtonia^ and the species are all western. So 

 little^howeYer^do the last three sections really differ from each other 

 that when new species have been discovered they have been placed 

 indiscriminately in the one or the other; and F.Mueller has in 

 several instances published them at once as BiBckeas, as Ilarmogias, 

 and as Babingtonias, giving his readers the option as to which 

 name to select. Of the additional genera proposed, Campliorosma^ 

 Schau., does not appear to me to differ in the slightest degree from 

 Harmogia^ and Tetrapora^ Schau., and. EricomyrttiSj Turcz., to be 

 inseparable from Bahingtonia. Tlie common BcEcJcea fruticosa of 

 the Indian archipelago is not exactly represented in Australia, 

 although _S. stenopliylla, P. MuelL^from Queensland, comes very 

 near to it. The three or four New- Caledonian species comprise 

 varieties of _B. virgata^ which is "widely spread in Eastern Aus- 

 tralia, and is there very variable, and a few forms nearly alUed 

 to it, but which may be constant enough to be maintained as 

 species. 



The following small genus, Astaktea, DC, is separated from 

 Bceckea by the purely artificial character of the stamens collected 

 in bundles or clusters, on which account it had always been placed 

 in a different subtribe,next to Melaleuca. This, however, appears 

 to be a mistaken view of real affinities. The clusters or bundles 

 of stamens in Melaleuca and its allies are always opposite to the 

 petals, whilst in Astartea tliey alternate with them, the arrange- 

 ment thus corresponding with that in those polyandrous species of 

 BiEcTcea in which the ring of stamens is interrupted opposite to the 

 petals. Indeed the Astartea amhigua^ F. MuelL, shows so nearly 

 the staminal arrangement of BcecJcea^ and the aspect of the whole 

 genus is so B(ScJcea'Y\kQ^ that it is kept up rather in deference to 

 general usage than from its intrinsic value. 



IlYPOCALTArMA, Endl., in its more numerous stamens, and 

 sometimes in inflorescence, assumes the aspect oi Leptospermumj 

 and it was first published as a section of that genus ; but the op- 

 posite leaves and the inflorescence of some species connect it with 

 Baeckeeje, where also it is more decidedly placed by the embryo, lij 

 indeed, it be constantly similar to that of the only species of which 

 the perfect seed has been examined. 



Balaustion, Hook., published also by Drummond under the 

 name of Cheynia^ has most of the floral characters as well as the 



