412 • BOTANICAL GAZETTE [December 



is directly above the pore in the perianth on the flat side of the 



flower. 



The perianth is never conspicuous either for its size or color. 

 It is greenish, slightly yellow, or brown. It contains on its corners 

 15-20 layers of cells. The cells remain parenchymatous at first, 

 but in the ripe seed they have rather thick walls and dense con- 

 tents. The cells bordering on the pore become increasingly 

 cutinized as development progresses. 



Earlier observations on this ovary are the following: Clarke 

 (5) describes the female flowers as consisting of three sepals formin 

 a tube not for the most part adhering to the ovary except at its 

 base and apex where it becomes trifid. He says : 



There is a peculiarity occurring in the calyx of Hedyosmum hirsutum or an 

 allied species (one of those in which the flowers are inclosed within thickened 

 bracts so compactly that the apex of the calyx and stigma are alone discernible) 

 which is probably quite singular: on removing bracteae, it is found that the 

 calyx does not completely cover the ovary, but has three large loopholes, 

 as it were, so that three flattened sides of the ovary are seen through it, although 

 it is quite continuous at the angles and crowns it with its three segments as in 

 other species. 



Solms speaks of a triangular ovary and of the perigon as 

 trifid at the apex, with the stigma sometimes alternate with the 

 lobes of the perigon. Baillon makes the following observation 

 concerning the perianth: " Moreover, the apex of the ovary bears 

 three short, thick, rounded wings alternating with its angles, 

 tw r o anterior and one posterior. Their morphological value is 

 still uncertain." Eichler speaks of a three-lobed half, or quite 

 superior perigon. Bentham and Hooker describe the tube 

 of the perianth as grown to the ovary. Van Tieghem says: 

 "La fleur femelle, egalement nue, se compose d'un seule carpelle 

 a style court renfermant un ovule orthotrope pendant; autour de 

 la base du style, la paroi de Tovaire se renfle quelquefois en trois 

 bosses epaisses (Hed.)." Engler describes briefly the female 



almost 



Clarke 



given a fuller description of the perianth, and each observer has 

 mentioned some of the facts. 



The question now remains as to the morphological nature of 

 the perianth. The occurrence of three vascular bundles in the 



