﻿^gg MR. J. MIERS ON THE LECYTHIDACEiE. 



Por the present, the consideration of true Barringtome^B will be deferred until the 

 LecytUdacea have been fully examined, premising that Gustavia must certainly Iq 



transferred to the latter group . 



In the LecyMdaceie there is no hypanthium in the sense above indicated ; hut the 

 very numerous stamens, always very small and short, are borne upon as many append- 

 ao-es, crowded upon a basal ring adnate upon a narrow flat annular epigynous disk ; and 

 this ring is always expanded into a peculiar development, called by Berg an urceohs, 

 hut which, by right, should bear the more appropriate name of an androphomm, long 

 before given to it by Mirbel \ This process is signally manifested in Gustavia, offering 

 an undeniable reason for excluding it from the BarringtoniecB, where it has been erro- 



neously placed. 



InGustavia (Plate XXXIII. a) the inflorescence, in its general features, resembles that of 



most LecythidaceiB ; its large yeUow flowers, upon pedicels, having in their middle 2 short 

 bractlets, possess a calyx adnate to the ovary, with a raised border, which is either short 

 and nearly entire, or is longer and divided into 6 acute sepals ; it has 6 or 8 very large 

 petals, unequal in size, imbricated in aestivation, severally inserted by their claws between 

 the disk and the base of the androphorum, where they are all agglutinated together, so 

 that, after fertilization of the ovary, the petals and androphorum, stiU united, faU off 

 together. The androphorum differs from that of most others in this family in being 

 equally expanded all round : it is very large, in the shape of an orbicular broad shallow 

 cup, and seemingly composed of several superposed plates agglutinated together, sue- 

 cessively smaller in diameter, aU alike deeply cleft on their broad free margins, into 

 numerous subulated segments, standing in many imbricated whorls which curve inwards 

 toward the style, the inner series very short, the rest gradually lengthening outwards, 

 until those of the outer series are nearly as long as the semidiameter of the cup. These 

 segments have generally been mistaken for the filaments of the stamens ; but in reality 

 they are not so, being subulated divisions of each plate, truncated or clavated at their 

 extremity, where the true filament is inserted, this being terete, very slender, and usually 

 shorter than the linear 2-celled anther, which at first opens by 2 apical pores. This 

 feature is well depicted by Berg \ The inferior ovary is divided into 6 cells, with many 

 ovules in each, severally suspended by distinct funicles from the placentary expansion 

 the central axis ; it is crowned by the sepals and annular disk, as before mentioned, the 

 vertex or space within the latter being flattish, with a short conical style in the centre. 

 The epigynous disk subsequently acquires a peculiar development in the frmt— a 

 spicuous feature throughout the whole family. The fruit is a globular pyxidium ahou 

 the size of an apple, flattened on its summit, where it is encircled by a band (the msx% 

 disk) enclosed within two parallel zones, the lower one formed by the vestiges o 

 decayed sepals ; the upper one denotes the inner margin of the enlarged disk, and o 

 the line of its opercular dehiscence : the pericarp is coriaceous, normally 6-ceUea, 



of 



' Mirbel, Ek'm. i. 2-40. 1 

 ■wtole systetfi of male organs 



rcecium, as opposed to gyncecium, is better 



mrlfir f.hp titlfi of andrOTohorum, wMle this 



the 



limited, as is here done, to that peculiar system of male organs found 

 * In Mart. Flor. Bras. fasc. IS. fah ^T^ flo- 14 



