STERCULIACE^E. Ig3 



spicuouslj acuminate. They resemble those of Eriodendron leian- 

 therum, DC, as figured by Cavanilles (under the name of Bornbax 

 erianthos), or those of E. Samamna of Martius, except that they are 

 for the most part abruptly acuminate. The petals are from 1? to 2 

 inches long, oblong-obovate, very tlvicldy clothed on the outside (except 

 their coalescent bases) ivith a silky wool, the hairs mostly reversed ; 

 their inner face glabrous, or a little pubescent towards the apex. 

 Filaments coriaceous, monadelphous only at the base, where they are 

 coalescent with the corolla, dividing almost as soon as they become 

 free into five phalanges, which answer to the five stamens of E. leian- 

 therum, E. Samamna, &c. These apparently alternate with the petals. 

 The phalanges, or united portion of the filaments, are half an inch or 

 more in length, linear-filiform, each dividing into 3 filaments, of about 

 the length of the petals, bearing single anfrachiose anthers. The 

 anthers, having all discharged their pollen, are too old to be minutely 

 investigated. They resemble those of E. Samauma, however, as 

 figured by Martius, are equally elongated, and apparently fixed by 

 the middle, and continuously one-celled (the cell of course longitu- 

 dinally bilocellate by the septum, here more or less persistent, which 

 normally divides the cavity, at least in an early state) ; but the turns 

 are closer and more heaped together, which is probably owing to their 

 condition after anthesis. They are very unlike those of E. leian- 

 tlierum; the corolla of which appears closely to resemble that of the 

 present plant, except in its greater size. The style is filiform, longer 

 than the stamens, and terminated by a capitate, obscurely five-lobed 

 stigma. 



I find no indication of any published species to which this plant 

 can belong.* Its remarkable peculiarity, it will be seen, consists 

 in the division (chorisis) of each of the five filaments into three; — - 

 notwithstanding which I do not hesitate to refer the species to the 

 genus Eiiodendron, modified by the needful extension of the generic 

 character. 



Chorisia speciosa, St. Hil., is mentioned by Dr. Pickering as found 



* In the Hookerian herbarium there is a specimen marked "Lima, ex Herb. Ruiz & 

 Pavon," and also "Loxa, 805," which appears to be the same as our plant, and may 

 therefore throw some light upon the habitat of the species. Unfortunately the stamens 

 are destroyed by insects, which renders perfect identification impossible. 



