OF SOME AMAZON TliEES. 11 



1737) has a tall straight trunk, almost hidden under a mass of 

 odoriferous flowers, and bearing at the apex a whorl of twice- or 

 thrice-forked branches. If the stem be continued beyond this, it is 

 by an innovation springing from below the whorl. In the same way, 

 Th. Cacao is proliferous, — the stems (for it is usually multicaul) 

 bearing an apical whorl of three dichotomous branches, then being 

 continued in an innovation, which, in its turn bears a similar 

 whorl, and so on. All this is very like what is above described as 

 occurring in Citrosma and Mabea, but corresponds more closely 

 with the branching of those species of Cordia, of which C. umbra- 

 culifera is the type. These Cordice have proliferous stems, and 

 bear at each successive apex a whorl of spreading (often drooping) 

 branches, which are 3-2-chotomously divided. In one Amazon 

 species the flowers grow on the naked trunk, in another they are 

 axillary on the ramuli, and in another terminal. In these two 

 latter forms, the whorl of branches, after flowering, decays and 

 usually falls along with the ripe fruit, leaving the stem naked and 

 with its successive innovations resembling sticks tied on to one 

 another at the ends. The chandelier- like appearance of the crown 

 of branches has induced the Venezuelans to give the name of 

 " Candeleros " to these Cordia. The French emigrants in Cayenne 

 call them, with almost equal propriety, " Arbres parasols." 



In some genera the branches are fascicled throughout, though 

 variable in number in each fascicle, often unequal in size, and not 

 springing from exactly the same point ; and yet their effect on the 

 habit and outline of the tree is the same as if the ramification were 

 regularly isochotomous. Ncea, Termmalia,and Bucida are branched 

 after this fashion. In such cases the main growth of a branch is 

 usually continued along the outermost ramulus of each fascicle ; so 

 that there is a continual recurvation of the secondary axes, quite 

 comparable to what is observed in the scorpioid cymes of some 

 Toumefortice, though less in degree. A most extraordinary form 

 is often assumed by Bucida angustifolia, DC, a tree abounding on 

 the sandy shore of the Rio Isegvo, where it exists for several months 

 in the year with barely its head out of water. Its crown is a wide 

 flat-topped reversed cone, and its short thick trunk puts forth 

 under water an almost equal-sized cone of radicles ; so that, when 

 the retiring waters leave it bare on the shore, it looks at a distance 

 like a gigantic hour-glass * ! 



* When I was at San Carlos del Eio Negro, and the famine grew sore in the 

 land of the Barres, as was hut too frequently the case, I was in the habit of 



