136 MR. G. BE^^XnAM OX MYKTACE^E. 



have been at various times pi'oposed as distinct genera. And, 



lastly, notwithstanding the facility of examining living specimens 

 of several sj)eeics common in our plant-houses, errors in some of 

 the important characters originally misunderstood have been 

 servilely copied by almost all modern botanists. 



After removing a few true Bseckeas, which the older authors 

 bad included in Leptospermitm^ and adopting as genera the sec- 

 tions Hi/pocaJymma, Endl., vrith a Ba^ckeoid foliage and probab]y 

 embryo, and Agonis^ DC, Avith Baeckeoid stamens and erect 

 ovules, both with the inflorescence of Melaleuca^ there remains a 

 not unnatural group, differing from Ktmzea and CalJistemon 

 chiefly in the stamens not exceeding the petals, and including the 

 genera Fabricia^ Maclclottia^ Jlomalospermum^ and Pericalymma^ 

 proposed or adopted by modern monographists. 



Fabricia was characterized and figured by Ga^rtner as having a 

 single winged peltate seed, filling each cell of the capsule ; but 

 this was a mistake. The ovules in both his species are exceed- 

 ingly numerous. It is true that, as in some other species of this 

 and allied genera, only one or two in each cell form perfect seeds, 

 the remainder either remaining small and abortive, or more or 

 less enlarguig into linear or misshapen barren seeds, and that in 

 the Fahricice the perfect ones are broad and acutely angular, or 

 more or less winged, whilst in typical Leptospermums the perfect 

 as well as the barren ones are narrow-linear ; but this appears to 

 be rather a sectional than a generic character. Ga^rtner's error 

 was owing to tlie fruits he examined in tlie Banksian collection 

 having been unripcj with the valves opened in dessieation, as fre- 

 quently happens in woody-fruited Myrtacea^. In this state the 

 unripe ovules, with the placenta, readily detach themselves In a 

 peltate mass, which Gtertner mistook for, and figured as, the seed, 

 without dissecting it. Gsertner's two species of Fabricia have, 

 moreover, usually ten cells to the ovary, whilst the typical Lep- 

 tosperrfiums have five or fewer ; but this distinction is not con- 

 stant. The Fabricia coriacea^ P. Muell., since reduced by that 

 author to a variety of Leptospermum {Fabricia) laevigatum^ has 

 from six to eight cells, and IlornaJospermum^ Schau., a single 

 species, w^ith the seeds and other characters of the section Fa^ 

 bricia, has always four cells to the ovary. 



The typical Leptosperma (section Euleptospermum\ with nu- 

 merous ovules and linear seeds, comprise a long series of forms, 

 which, according to Schauer, or to the earlier views of Mueller, as 

 published by Miquel, or to several horticultural botanists, would 



