Mr. J. Miers on some genera of the Icacinacese. 113 



lodes ; but fully acknowledging all that is there affirmed, I may 

 venture to show, that a yet stronger and much closer extent 

 of analogy will be found to exist in the structure and deve- 

 lopment of the floral parts, as well as a greater approximation 

 in habit, to what we find in Stemonurus and Pennaniia. From 

 the facts shown below, it will be seen that Sarcostigma accords 

 with nearly all the essential characters I have endeavoured to 

 establish in the preceding series of memoirs, as the leading fea- 

 tures of the IcacinacecBj viz. trees with alternate, glabrous, coria- 

 ceous, petiolated, exstipulate leaves ; an axillary racemose inflo- 

 rescence, with small flowers, more or less polygamous, and di- 

 stinctly articulated on a short pedicel ; a small cupshaped, per- 

 sistent calyx supporting the fruit, and unchanging with its 

 growth ; a corolla of four or five fleshy, linear petals, with val- 

 vate aestivation, arising from the hypogynous or stipitated sup- 

 port of the ovarium; free stamens, equal in number to, and 

 alternate with the petals ; introrse 2-lobed anthers ; an ovarium 

 presenting a similar form, the same internal structure, and the 

 subsequent development of that seen in Stemonurus and Pen- 

 nantia, and a fruit, in all appearance, closely analogous to that 

 existing in those genera. Hence it seems evident from the facts 

 here shown, that wherever Pennantia, Stemonurus^ and Platea 

 are placed in the system, Sarcostigma should follow in juxta- 

 position with them, unless the evidence now wanting, of the 

 structure of its seed, should tend to a diff'erent location. If 

 therefore Sarcostigma be found to hold a relation with the Phy- 

 tocrenecBy the questions will naturally arise, whether this hitherto 

 dubious family should not be brought into a more proximate 

 position in the system with the IcacinacecBj or whether I have 

 been in error in referring the genus under consideration to the 

 latter family. The group of the Phytocrenea was first proposed 

 by Endlicher as a suborder of the Menispermacece, a family with 

 w*hich they hold little relationship. Prof. DeCaisne, if I mistake 

 not, first pointed out the identity of Phytocrene with the Gyno- 

 cephala of Blume, a genus placed among the Artocarpacece : hence 

 Phytocrene and Nansiatum were removed by Prof. Lindley and 

 other botanists to that family. This conclusion appears to me 

 to have been too hastily drawn, for the Artocarpacece differ from 

 them essentially in their stipular leaves, the presence of only a 

 single floral envelope, which is often imperfect or altogether 

 wanting, in their having fewer stamens than the number of the 

 lobes of its perianthium, in their bifid style, which is often 

 basilar, in their ovarium, with only a single suspended ovule, 

 which is amphitropal or orthotropal, and an exalbuminous seed, 

 often erect, though sometimes pendulous, with a thickened testa, 

 and thick, fleshy cotyledons, often unequal in size. Phytocrene is 

 Ann.^ Mag. N. Hist. Ser. 2. Vol.Ji. 8 



