66 MR. bentham's synopsis of leonotideje, 



opposite leaves with interpetiolar stipules, their valvate calyx, and 

 in many respects in the general arrangement of the parts of the 

 flower ; but differ in inflorescence, in the aestivation of the petals, 

 in the more complete consolidation of the styles (although in 

 the latter respect they agree with Broussaisia) , and in the pen- 

 dulous collateral or clustered ovules. With Lythracecs they agree 

 more generally in habit and inflorescence, in the aestivation of the 

 calyx, and completely consolidated styles ; and where the ovary is 

 free, it is often enclosed in the base of the calyx, as in Lythracece ; 

 they differ, however, in their dotted leaves, interpetiolar stipules, in . 

 the aestivation of the petals and insertion of the ovules, and in 

 their albuminous seeds. The arrangement of the petals and 

 stamens is also somewhat different, even in the genera which have 

 the ovary free, as in Lythracece, In the latter order, the disk which 

 bears the petals and stamens is so completely combined with the 

 calyx-tube as scarcely to thicken it, or form any perceptible border 

 at its edge, where the petals are inserted close among the calyx- 

 teeth, and the stamens are usually inserted more or less irregularly 

 lower down on the inner face of the disk. In Legnotidew, the disk, 

 whether very short in the base of the calyx-tube, or lining the 

 tube up to the teeth or lobes, is always very perceptible, and 

 sometimes thick and fleshy ; its margin is usually more or less pro- 

 minent, either as an entire or crenate ring, or as a short cup-shaped 

 tube round the ovary ; the stamens are usually inserted either on 

 the margin, or more frequently just below the margin on the 

 outside, and the petals outside the stamens at the base of the disk, 

 that is, in the angle which the free part of the disk makes with 

 the calyx-tube. It is probably this disk which induced Dr. Wallich 

 originally to consider the GynotrocJies as a species of Microtropis 

 among Celastrinece. 



The fringed and folded petals of most Legnotidece have perhaps 

 some analogy to the crumpled petals of Lythracece ; the manner 

 in which they are folded over the larger stamens, or over all the 

 anthers, has some analogy to the hooded petals of Mliamnacece, 

 which enclose the anthers in the bud; but there is very little 

 other affinity with that order. So the opposite dotted leaves or 

 the inflorescence give a general resemblance, without much affinity, 

 to MyrtacecB and Memecylew. 



The fringed petals of Gardner's Anstrutheria had induced him 

 to refer that plant to Elceocarpew ; but that and the valvate calyx 

 are almost the only characters in which the Legnotidece coincide 

 with that tribe ; still less can I see any affinity between Legnotidece 



