132 MR. G. BENTITAM OX MYIITACE.E. 



combined as give greater facilities for the general arrangement of 

 the Order if treated as generic than as sectional. 



Tribe II. Leptospeume^e. 



The long series of 2- or more-celled capsular genera forming 

 the tribe Leptospermese are nearly all Australian, and most of 

 them exclusively so, or at most emitting a few species into New 

 Zealand, New Caledonia, or some of the islands of the Indian 

 archipelago and the Pacific ; and most of them, although perhaps 

 not quite so vague as the baccate Myrtea?, yet pass into each 

 other through intermediate species which render their delimita- 

 tion much more difl&cult than in the case of Chamselauciese. On 

 the other hand, some of the diiferential characters are so striking, 

 thatjif the ambiguous species were overlooked, there would appear 

 to be several well-marked subtribes. The first one especially, the 

 Bseckeese, has always, as far as known, the minute cotyledons of 

 Chamselauciese, accompanied usually by opposite leaves and defi- 

 nite or uniseriate stamens ; whilst in the following subtribes the 

 cotyledons are as long as or longer than the radicle, the stamens 

 usually indefinite, and 2- or pluri-seriate, with varying leaves. The 

 apparently positive character of the embryo remains, however, to 

 be confirmed in some intermediate species, of which the ripe seed 

 is unknown. 



ScnoLTZiA, Schau., which I have placed first, as connecting 

 Micromyrtus with BcscJcea^ is so near the latter genus as to have 

 been made a section of it by De Candolle. It also shows both 

 the forms of anthers prevalent in ^ceckea. As, however, the ar- 

 rangement of the ovules, 2 superposed in each cell, is accompanied 

 usually by a rather peculiar inflorescence, and as some species, by 

 their apparently incomplete dissepiments of the ovary, show an 

 approach to Micromyrtus ^\t has been thought convenient to main- 

 tain the genus as distinct. Piptaiidra, Turcz., has no character 

 ■whatever to distinguish it from other species of 8clioltzia^ which 

 have constantly, or occasionally, 3 cells to the ovary. 



The typical genus, B^eckea, Linn., was originally supposed to 

 be well characterized by definite stamens and a 2- or more-celled 

 ovary with a heath-like habit ; but as numerous exceptions have 

 been successively discovered, it has been variously understood by 

 different botanists. Some include ThryptomenCy Micromyrtus, 

 Scholfzia, and Astartea, all of which have a similar habit, but with 



structural differences, which may justify their maintenance as 



