including a new Airangement of Phanerogamous Plants. 199 



of organization, one of them conterminous with or passing into 

 Endogens, and the other the most remote from them, differing 

 from each other more in the position of the carpel when single 

 than in any other character. And as placentse are prolongations 

 from one common pith, and as ovules are analogous to buds, in 

 connection with which the wood is formed, it becomes a question 

 whether there is not an analogy between the formation of wood 

 externally on the stem, and the development of ovules and car- 

 pel on the external or anterior side of the flower, especially in 

 such forms of inflorescence as spikes, racemes, &c. If so, it 

 would be a reason for regarding those Orders in which the single 

 carpel is anterior, as having a more perfectly exogenous cha- 

 racter, the first ovules and carpel being constantly formed on the 

 external or anterior side, showing more tendency to exogenous 

 structure than where they are produced irregularly or posteriorly. 

 Consistently with such an analogy, stems imperfectly exoge- 

 nous are confined to the Heterocarpous Division ; and it may be 

 observed also, that in Endogens the first leaf of an axillary bud is 

 for the most part lateral or posterior, and that the succulent leaf 

 forming the clove of Allium, although not the first leaf, is pos- 

 terior or occasionally lateral, while among Exogens the first leaf 

 is commonly lateral or anterior, of which latter Xylophylla and 

 Phyllanthus are remarkable instances*. And finally, this ante- 

 rior or external development sometimes extends to the floral en- 

 velopes, as in Acanthus spinosa the calyx and corolla are deficient 

 posteriorly ia common with the stamens and (in Mendozia) the 

 ovary of Acanthacese, and a comparison may be made with the 

 modes of growth and branching of the axis. 



Note on Table III. 



The main purpose of this Table is to show that the arrange- 

 ment adopted in Tables I. and II. is not inconsistent with 

 well-established affinities. The Subdivisions therefore stand in 

 the same succession to each other as ia Tables I. and II., by 

 which the epigynous Alliances are now brought into relation 

 with each other, and most of the Natural Orders become so 

 placed as to be within the range of their more immediate affinities ; 

 and, indeed, by contracting or widening the separated portions 

 of the Subdivisions ad libitum, but few of them would remain 

 unassociated with their nearest allies. To this, the Leguminous 



* The leaves of Xylophylla are regarded as true leaves, because — 1. The 

 woody circle in the petiole is incomplete in its upper part, a character 

 common in petioles, and which has been termed " the horse-shoe mark." 

 2. The venation, viz. the branches terminating in the depressions of the 

 serratures, is frequent in Mercurialis annua. 3. The scales at the bases of 

 the leaflets occur also in Schottia. 



