224 PROCEEDINGS OP THE AMERICAN ACADEMY 



describes the seed as destitute of albumen, the cotyledons as thick and 

 plano-convex. Torrey, correct as far as it goes, wrote : " Embryo 

 curved in thin fleshy albumen." I find the embryo curved almost 

 into a circle, and a very thin stratum of firm albumen, which might 

 be mistaken for a thick tegmen lining the crustaceous testa. What 

 has not been noticed is, that the oval and not very thick cotyledons 

 are longitudinally convolute, in the manner of Cadellia, F. Miiller, 

 as figured by him in Fragm. Phyt. ii. t. 12. As Bentham, in Fl. 

 Austral., describes a monocarpellary species of this genus, with basal 

 style, and which has dotless leaves, it seems that this genus is even 

 more than Gneorum allied to Cneoridium, and that the technical char- 

 acter of the order vanishes. 



CHOISYA, HBK. The gynaecium is quite incorrectly figured 

 and described by Kunth. It is in DC. Prodr. i. 724 that the correct 

 character "capsida b-rostrata" first appears; and the next year (1825) 

 Adrien Jussieu, who well describes the pistil, appends to his account, 

 " Fructus (teste Bonpland) capsularis, 5-sulcus, h-rostratus" Finally, 

 Baillon (Hist. PI. iv. 471) adds, " Cocci 5, bivalves, endocarpio 

 soluto " ; so he must have found at Paris dehiscent fruit, from which 

 the seeds had fallen, for these are still undescribed. I can now add a 

 second species to this genus, and describe the seeds ; viz. G. dumosa, 

 Astrophyllam dumosum, Torr. Pacif. R. Rep. ii. 161, and Bot. Mex. 

 Bound. 42, of which we now possess good flowers and fruit, which 

 quite accord with Choisya. The seeds (solitary or rarely in pairs) 

 are reniform, with nearly smooth subcrustaceous testa, and arcuate em- 

 bryo in thin albumen. There is either a deciduous caruncle, or else 

 a small and definite portion, of the thin-cartilaginous endocarp falls 

 away attached to the hilum. 



PTELEA. P. pentandra of Benth. PI. Hartw. is apparently 

 Rhus Toxicodendron. P. aptera, Parry, of Lower California, is very 

 remarkable for its nucumentaceous and turgid fruit surrounded by a 

 very narrow wing or else quite wingless. As Planchon and Triana 

 have stated, De Candolle was quite wrong in adducing Amyris ele- 

 mifera, L., to P. trifoliata. Catesby's figure is evidently that of an 

 Amyris, probably the small-leaved form of A. maritima, Jacq. (var. 

 angusiifolia) , which is found on the coast of Florida, although the 

 leaflets are represented as too broad and rounded at base. The 

 habitat " Carolina " is a mistake, and it has no foundation in Catesby's 

 account. 



