8 ME. E. SPRUCE ON THE MODE OF BRAtfCBIXO 



the two groups are so decided, that, should the lew polyandrous 

 Diospyri found in other parts of the world possess the same verti- 

 cillate ramification as the Amazon species, I should be disposed to 

 place them in a genus distinct from Diospyros. 



Macreightia myristicoidea, PI. Am. (Uaupes), another Ebenaceous 

 plant, has the branches in whorls of three, and has a very similar 

 habit to that of Myristica debilis, which grows along with it. 



Many Artocarpeae show a tendency to a whorled ramification. 

 In Cecrqpia the normal ramification seems to be thus : branches 

 in fives, each branch ending in three ramuli, or, if the division is 

 carried further, the innermost ramulus is the main one ; so that 

 the branches are whorled, and then trichotomous. There is some- 

 times a suppression of one, two, or three branches in the primary 

 whorls, and of one of the ramuli in the fascicles : and very often 

 the branches of a whorl do not spring exactly at the same height 

 on the stem ; but as they are always within a few inches of each 

 other, while the whorls are several feet apart, the verticillate cha- 

 racter is distinctly preserved. 



Having considered some of the more remarkable cases of pyra- 

 midal or pai*aboloidal outline, where the crown of a tree from a 

 broad base tapers gradually to the summit, let us consider the 

 opposite form, where the crown has a broad flat or subconvex 

 summit, and tapers gradually downwards to the insertion of the 

 lowest branch, i. e. approaches to obconieal or obpyramidal. The 

 most perfect examples of this mode are found in trees with oppo- 

 site leaves and terminal inflorescence ; and there is none more 

 notable on the Amazon than the Mulatto-tree (EnkylistaSpruceana, 

 Benth.), a tree 80 to 100 feet high, frequent all along the banks, 

 and whose polished brown trunk and narrow obconieal top mark it 

 out to view among the adjacent trees. If we examine young spe- 

 cimens of this tree, or of many Psychotrice, Cephaelides, &c., we 

 shall find the first flowers to be borne in a cyme at the apex of the 

 primary axis. Prom beneath the cyme proceed two opposite and 

 equal branches, which, in their turn, bear each a cyme at the apex 

 and a pair of branches below it. Thus a regularly dichotomous 

 ramification is generated, and the growth of the tree is continued 

 along a number of equal independent axes, whose number is being 

 continually augmented ; for observation enables us to lay down 

 this axiom : that when any axis, whether principal or partial, flowers 

 at its apex, its upward growth is completed; it may increase in bulk 

 by the addition of tissue to its length and breadth, but not by pro- 

 longation from an apical bud. Exceptions to this law are only so 



