MK. G. BENTIIAM OjV MYKTACE^. 137 



constitute between thirty-five and fortj'-, or even more, species ; but 

 tbeir characters are exceedingly vague and variable, and they 

 appear all to be connected by an almost infinite series of inter- 

 mediate gradations. In arranging the large mass of specimens I 

 had before me, for the ' Elora Aiistraliensis,' 1 thought I could 

 distinguish twelve types which might rank as species, besides 

 two rather more distinct ones with 3-celled ovaries ; but it must 

 be confessed that the characters of those twelve are often unsa- 

 tisfactory, and a much furtlier reduction, as now proposed by 

 r. Mueller, might well be justified, even though it might lead to 

 the uniting tlie whole twelve into a single one. 



The Le^ptosperma from the Indian archipelago belong to the 

 same series as the twelve Australian ones ; and one of them, L. 

 amloinense^ DC, appears indeed to be identical with the Austra- 

 lian L.Jlavescens^ Sm. They are proposed by Korthals as a dis- 

 tinct genus under the name of Macklottia. He does not say 

 upon what grounds ; for he gives no amended character of Lepto- 

 sper^imim ; but from that of his MacJclottia it is probable that he 

 founded it on the imbricate calyx-lobes, previous authors having 

 described them as valvate, an error adopted by Blume in his de- 

 tailed character of the archipelago species ; and even Miquel 

 (Fl. Ind. Bat.i. pars 1. p. 403), in reducing MacJclottia, says, " Ca- 

 lycis . . . limbi . . . laciniis sestivatione baud imbricatis." I have, 

 however, in both the archipelago species, as well as in all others, 

 found the calyx-lobes constantly imbricate in the bud, as described 

 b}? Korthals. 



.The three remaining Leptosperma^ all from South-west Aus- 

 tralia, form the section Pericali/mma^ Endl., characterized by a 

 reduced number of ovules and a peculiar habit. Schauer, in rais- 

 ing the section to generic value, includes in the characters the 

 3-celled ovary ; but that, as has been seen, occurs in two species 

 of Uuleptospermum with numerous ovules. As, moreover, the 

 three species of Pericalymma are not very definitely distinguished, 

 it may turn out that they form rather a single species of Eulepio- 

 spermum than a sectional group. 



KuNZEA., Reichb., differs from Leptospermum in Its exserted 

 stamens, on which account the species foi'merly known were re- 

 ferred to Metrosideros \ but the foliage, inflorescence, and structure 

 of the ovary are entirely as in Leptospermum^ and some of the 

 smaller-flowered species have the stamens sometimes so little 

 exserted as almost to pass into that genus. Trom Callistemori 

 Kunzea diflers usually in the shorter stamens, the capitate, not 



