X68 MONADELPHIA DODECANDRIA. Bomhax. 



smooth. Border in fveneral obscurely three or four-parted ; 

 as the corol expands, these are often so much split, as to ap- 

 pear four, five, or more cleft, ssepa rating- to near the perma- 

 nent base, which, with the corol and stamina, drop off' in one 

 body. Corol of five, very distinct, oblong, first spreading, 

 then recurved, contorted, smooth, deep red, fleshy, oblong- 

 petals, of about twice the length of the stamens. Filaments 

 in two series ; the intei ior ones pressing on the style, and 

 consisting of five longer and thicker, and ten shorter; the 

 exterior series contains from fifty to sixty, united into five, 

 distinct phalanges, all these are united at the base, into one 

 fleshy envelope for the germ. Anthers incumbent, involute, 

 reniform, those of tlie larger filaments of the interior series are 

 double, on all the rest single pollen, the colour of Scotch snuflT. 

 Germ conical. Stifle longer than the stamina. Stiffmas five, 

 subulate, recurved. Capsule oblong, tapering equally to- 

 ward each end, five-celled, tive-valved, downy on the outside. 

 Seeds numerous, obovate,smooth, except a sharp, cross-shaped 

 ridge on one side, immersed in a very large quatility of very 

 fine silky wool ; this wool does not adheie to the seeds, but 

 rather seems to grow from the inside of the valves of the 

 capsule. 



In my description of this tree, I have been particularly 

 full, because I am inclined to think it ditfers essentially from 

 the West Indian species, Miiicli Linnaeus called Ceiba. 

 Jacquin's full, and I presume, accurate description of that tree 

 is now before me, and leaves in my mind little doubt of their 

 "being very difterent. Tins and the iornier pent a ndr urn, are 

 the oidy species of this genus 1 have hitherto met with in In- 

 dia, except B. f/ossjjpinum, Avhich is polyandrous. If this 

 be not what Linnneus calls B. heptaphyllum, and Jacquin, 

 Septcnatum, I cannot say what it is. He quotes Moul elavou 

 of the Hortus JMalabaricus for it; and at the same time 

 points it out to be an American tree. Be that as it may, 

 there can be no doubt of Rheed's i)lant being the one I have 

 MOW described and fiijured. 



