Liriosma, Leretia, Poraqueiba, and Emmotum. 359 



wedges between the radiating spaces j near the summit these 

 spaces terminated in one point, and in the solid part above this 

 I discovered with some difficulty a very minute embryo, about 

 half a line in diameter, consisting of two distinct, orbicular, flat, 

 foliaceous cotyledons, with a notch on the margin for the inser- 

 tion of the radicle, which was unfortunately lost in its extraction ; 

 but the polished indentation of the entire embryo, left in the 

 substance of the albumen, indicated its shape and relative size 

 in regard to the cotyledons. The evidence of this structure is 

 therefore nearly complete. In the diminutive size of the em- 

 bryo, imbedded in the summit of the albumen, this structure 

 quite agrees with that of Heisteria and other genera of the Ola- 

 cacece ; and it may be further remarked that I have observed in 

 the albumen of Heisteria and Liriosma a constant large vacant 

 space, running from its base up its centre, and terminating in 

 the more solid part below the summit, in which the small em- 

 bryo is imbedded. 



From the structure of the ovary, fruit, and seed, combined 

 with other characters, it is manifest that Aptandra is nearly 

 allied to the Olacacece, agreeing with that order in the form and 

 aestivation of the calyx and petals. The singular enlargement of 

 the calyx, which ultimately half envelopes the fruit, occurs in a 

 very similar manner in Olax, Heisteria, and some other genera ; 

 but Aptandra is very distinct from every genus of that family in the 

 remarkable confluence of the stamens into a long, thick, mona- 

 delphous tube, which embraces the style, and bears a number of 

 extrorse anther-cells, adnate below its summit, just as in Canella, 

 but which burst by the deflection of their outer valves, as in 

 Diclidanthera. 



This striking peculiarity might well claim the right of Ap- 

 tandra to be the type of a distinct family allied to Olacacece; but 

 I will not venture to propose it until other analogous genera 

 are discovered : in the meantime it may remain as a suborder of 

 that family, bearing the name of Aptandrete. 



To the generic characters of Aptandra, as before given {loc. 

 cit. p. 201), we may therefore now add : — 



Drupa magna, sicca, globosa, calyce persistente aucto cupulari 

 laxe semicincta, pericarpio coriaceo indehiscente, 1-locularis, 

 1-sperma. Semen imo loculi affixum; integumenta ignota; 

 albumen copiosum, carnosum, a basi ultra medium radiatim 

 excavatum ; embryo parvus, in illo summum versus sepultus, 

 cotyledonibus orbicularibus, valde foliaceis, radicula brevi te- 

 reti supera 4-plo longioribus. 



1. Aptandra Spruceana, loc. cit. vii. 202; Contrib. Bot. i. 3. 

 tab. 1. 



