40 DIMORPHIC BRANCHES IN TROPICAL CROP PLANTS. 



RELATION or DIMORPHIC BRANCHES TO HABITS OF GROWTH. 



Other cacao trees, both wild and cultivated, fail to show these 

 habits of growth. Instead of the erect main stem, with branches in 

 rosette-like clusters, the trunk divides near the ground into many 

 oblique arms that form a broad spreading top of dense foliage, 

 entirely unlike the open, irregularly distributed foliage of the trees 

 with tall upright trunks. Planters of cacao have recognized cultural 

 differences betAveen the two forms of trees, the low, spreading type 

 being preferable for jjlantation purposes to the tall type with the 

 whorled branches. 



It has been supposed that the different habits -of growth betoken 

 two different varieties of cacao, but seedlings from the spreading 

 trees have not been found to show any tendency to reproduce the 

 spreading habit of growth. If the spreading trees had any other 

 character in common, the idea of a varietal difference might still 

 appear to have some justification, but the fact is that lioth kinds of 

 trees show the same general range of individual differences in the 

 characters of the fruits, which are the only parts of the plant that 

 lend themselves to careful comparison. The serious difference lies 

 in the fertility, for the low, compact trees that shade their own short 

 trunks and the ground underneath them appear to thrive much bet- 

 ter in plantations than trees of the other type, and bear larger crops. 

 In eastern Guatemala, where this matter was studied in some detail, 

 it was the opinion of a very intelligent cacao planter, Don Eicardo 

 Fickert-Forst, owner of the Trece iVguas estate, that the low, spread- 

 ing trees would bear, on the average, at least twice as much cacao as 

 the others, and that they would continue to be fruitful for a longer 

 period of years. Efforts had been made to obtain more of the 

 spreading trees by planting seeds from trees of this form. 



The failure of such attempts can be explained after the serious 

 differences between the two kinds of branches are recognized. The 

 low, spreading trees have this desirable form because they do not 

 produce any of the upright shoots and whorls of branches. Their 

 method of branching is the same as that shown on whorl branches, 

 that are incapable of forming uprights, as already explained. Al- 

 though there is no indication of a whorled arrangement of the main 

 branches of the spreading trees, it may nevertheless be considered 

 that the tops of these trees represent the development of only one or 

 two of the branches of an original whorl, and this would afford an 

 adequate explanation of the formation of a different type of tree. 



The inability of the whorled branches to produce any upright 

 shoots Avould explain why a tree top formed from such a branch 

 would not have any of the strong upright shoots, but would produce 



198 



