flowers on tte same plant. The B. Malaharicay 

 Roxb., seems to be a species quite distinct from Dry- 

 ander's, and Rheede's plant, but it is, like my plant, a 

 Diploclinium^ apparently more nearly allied to it than 

 to Rheede's. 



In addition to these, Professor Nees, in his ex- 

 cellent and most elaborate monograph of the order 

 (Systema Laurinarum), has availed himself, for group- 

 ing his genera into tribes, of the duration of the leaves, 

 whether deciduous or evergreens — of the inflorescence, 



In these characters I, in referring to the lobes of ^hether umbeUed or panicled-of the dehiscence of 



the perianth, have followed my predecessors in call- 

 ing them petals, which is not quite correct though so 

 thoroughly petaloid in texture and appearance. 



Laurace^. 



The formation of the flowers in this order being 

 somewhat peculiar, I have in several of the follow- 

 ing plates endeavoured to exhibit their distinguishing 

 featmres by means of diagrams. A few introductory 

 remarks explanatory of these diagrams and of the 

 parts they are intended to represent, seem necessary. 



In this order the flower is nferior, usually bisexual, 

 with a six-lobed perianth, 12 more or less perfect 

 stamens, and a 1 -celled ovary with a single pendu- 

 lous ovule. The lobes of the perianth form a double 

 row or series 3 and 3, or rarely two or four in each. 

 Each of these lobes has in front, 2 stamens, forming 

 together 4 rows of 3 each : those of the 2 outer rows, 

 next the perianth, are usually perfect with the anthers 

 opening inwards, those of the third row opposite the 

 first or outer are also usually perfect, but differing 

 from the preceding in having two pediceUed glands 

 at the base of the filament, and then' anthers opening 

 outwards. Those of the inner of 4th row, opposite the 

 2d, are rarely perfect, being usually antherless fila- 

 ments, or what are called stamijiodes. In the follow- 

 ing diagrams the stamens of the outer rows being 

 normal (a filament and perfect anther) their places are 

 marked by a small o, those having gland-bearing fila- 

 ments by a double circle oo ; and lastly the staminodes 

 by a point. In some genera the anthers of both the 

 interior rows are perfect and glanduliferous, the dia- 

 grams show these by the increased number of double 

 00. In some the inner row is altogether wanting, these 

 are equally shown by the absence of points. These 

 differences are employed as generic characters. Some 

 genera have 1 -sexual flowers; the analyses show 

 these by representing separately the male and female 

 flowers when both were procurable. Others have 

 several flowers aggregated within an involucrum, 



forming a head or simple umbel. Cylicodaphne tet- 



ranthera^ &c., furnish examples of this arrangement. 



L 



In this order the anthers are 2- or 4-celled, not, as 

 in other families, opening by slits or pores, but by 

 valves which separate from below, and turn back 

 towards the apex as shown in aU the plates. 



The ovary is superior and free, except in a few 

 genera where it is more or less completely embraced 

 by the tubular base of the perianth. (See Cryp- 

 tocarya.) 



the anthers, whether opening at the apex or below 

 the apex — of the fruit, whether free or more or less 

 inclosed within the perianth — the stammodes of the 

 4th series, whether wanting, imperfect, or distinctly 

 3-angular — the limb of the perianth, whether persis- 

 tent or deciduous — in the former case, whether har- 

 dening into a cup or not hardening : and in the latter 

 whether the bases of the lobes are persistent and 

 truncated or altogether deciduous from the tube. 

 The leaves are also used in the limitation of these 

 groups, whether, namely, they are triplinerved or 

 penninerved, and reticulated. In Cinnamomum^ they 

 are 3-nerved or triplinerved: generally less dis- 

 tinctly so in the followuig. To show how these char- 

 acters are used, I shall introduce verbatim Nees' 

 Synopsis or Key to the Tribes, '* Clavis Tribuum," in 

 which he exhibits in a tabular form, their applica- 

 tion in practice. 



The characters employed for the limitation of the 

 genera are sometimes very minute, and their value 

 at first sight appai-ently so inappreciable, that I re- 

 peatedly thought, when I first entered on the study 

 of the order, that sub-division had been carried to an 

 unnecessary degree of refinement, an opinion which 

 increase of knowledge, though it has not altogether 

 removed, has certainly not strengthened, but it is one 

 on which, considering my limited opportunities of 

 studying the order as a whole, it would not have been 

 safe to act. I have therefore as much as possible 

 availed myself of named specunens for representa- 

 tion, and have only in two instances altered names 

 given by Nees, but hope that in both cases the learn- 

 ed Professor wiU adopt the alterations. 



Before concluding these introductory remarks, it 

 may be well to advert briefly to an organ, if such it 

 may be called, which, if I rightly understand, seems to 



have given the Professor some trouble ; I allude to 

 the pediceUed capitate glands of Tetranthera, These* 

 bodies he at different times calls both glands and 

 staminodes, a most inconvenient confusion of terms. 

 For myself I can see no reason for considering them 

 any thing else than a modified form of the glands 

 found in every genus in the order* But at the same 

 time I look upon the modification as so peculiar, and 

 of such rare occurrence, that I think it might have 

 generic value attached. They exist in several, but 



not in all the species of Tetranthera^ those in which 



they are present, at least so far as my experience ex- 

 tends, seem to me to form a distinct and well-mark- 

 ed genus, which might be beneficially separated from 

 the rest of the genus. But to this I shall advert 

 more at large by and bye. 



( 10 ) 



• 



