30 CENTRAL AMERICAN RUBBER TREE. 



CASTILLA NOT A GENUINE FOREST TREE. 



The native population of the Central American region is commonly 

 supposed to have been much more numerous previous to the Spanish 

 conquest, and the numerous and widely distributed ruins prove the 

 former existence of relatively civilized communities in localities which 

 even in the time of Cortez were apparently forgotten and overgrown 

 with forests as they are to-day. But notwithstanding the former 

 civilization of these regions, there seems not to have been found 

 anywhere in Central America an indication of permanent agriculture, 

 such as terraces, walls, or irrigating ditches. The agriculture of the 

 ancient Indians was probably like that of the modern, in that each 

 head of a family cut down and burned each year a new piece of forest 

 to plant his farm or " milpa." Where the population is large and 

 old forest is no longer accessible the second and successive growths 

 are cut at intervals of a few years until the tropical rains have washed 

 away all the fertile surface soil and the district becomes, for the time, 

 a desert, and is afjandoned by its human inhabitants. Such deserted 

 country is covered first by a coarse grass and then by a scattering 

 growth of pines, which are in turn crowded out by an invasion of 

 tropical forest vegetation, at first in the more sheltered and humid 

 ravines and valleys and then over the whole area. At low elevations 

 the trumpet tree and Castilla form a part of the vanguard of the new 

 growth, and the Attalea palm is its most striking species. But it is 

 only a question of enough time for these and their accompanying 

 species to be overcome and well-nigh exterminated by what may be 

 termed the permanent forest. 



When one sees the Indians of to-day clearing, burning, and planting 

 precipitous and scarcely accessible cliffs it becomes easy to believe 

 that little fertile land in Central America, if any. is occupied by truly 

 primeval forest, and easy also to understand that the abundance and 

 wide distribution of Castilla may depend upon human activity even 

 more than upon natural agencies. Arguments based upon the assump- 

 tion that Castilla is a genuine forest tree may accordingly be dis- 

 missed as of little agricultural significance. 



Mr. O. H. Harrison, manager of the rubber estate at La Zacualpa, 

 was much interested in this view of the place of Castilla in nature, 

 because he had already noticed that clusters of wild Castilla are met 

 with in the forest only where some natural or artificial clearing had 

 been made. Moreover, an examination of the literature of rubber 

 shows that the facts are not new. though their significance has been 

 concealed by the explanation which accompanies the following original 

 account of the details learned from the rubber gatherers of Nicaragua: 



The trees prefer humid and warm soils, hut not marshy, clayey, or gravelly ground, 

 and the presence of these trees is looked upon as an indication of a fertile soil. It is 

 not distributed irregularly through the forests, but sometimes in little groups, more 



