IMPROVEMENT BY SELECTION. 31 



or less isolated, such a group being termed a mancha (spot). This grouping is the 

 normal state, and is believed to be caused by monkeys dropping the seeds near an 

 isolated tree, as they are very fond of the pulp by which the seeds are surrounded. 

 The trees are distributed in vetas (veins) or bands, either in a north-to-south or east- 

 to-west direction, the first probably caused by monkeys, by the trees being on a decliv- 

 ity, or by water, and the second by the wind, which daily blows in that direction. 

 This irregular distribution has led M. Levy to the opinion that in cultivation they 

 should be interspersed between other trees rather than form separate plantations, as 

 he thinks that this sympathetic and antipathetic tendency should not be lost sight 

 of. The hule is often near water courses, and nearly always on the banks. Trees of 

 small groups give a better net produce than those composing large groups. a 



From the scientific standpoint these explanations appear quite inade- 

 quate, since the causes which they suggest are those which are in con- 

 tinuous operation, and if effective in spreading Castilla at the expense 

 of other forest trees would have given it a general preponderance long 

 since. All the facts are, however, comprehensible on the supposition 

 that the growth of Castilla depends upon opportunities which are rela- 

 tively infrequent in undisturbed forests, as compared with regions 

 inhabited by the Indians and subject to their primitive agriculture. 



IMPROVEMENT OF RUBBER TREES BY SELECTION. 



Instead of being able to dispense with agricultural knowledge, skill, 

 and caution, the rubber planter needs an extra supply of these, since, 

 without the advantage of adequate experience, he has the added 

 responsibility of choosing favorable natural conditions, applying cor- 

 rect cultural methods, and securing the plants most suitable for the 

 circumstances under which he must operate. That American planters 

 have given their attention so exclusively to Castilla, and those of the 

 East Indies to Hevea, is not the result of any demonstration of the 

 cultural superiority of the one tree or the other, and the desirability 

 of man}^ other species reported as promising remains to be determined. 

 It is entirely possible that no one species will be found to have a supe- 

 rior value under all conditions and be planted to the exclusion of all 

 others. Rubber, like starch, is produced in nature under varied con- 

 ditions, ranging between deserts and swamps. The number of culti- 

 vated rubber plants will probably never equal that of the starchy 

 cereals and root crops; but there is the same practical reason why the 

 cultural requirements, hardiness, vigor, and productiveness of the 

 different rubber plants should be considered, and not merely those of 

 the distinct genera and species, but those also of the differing varie- 

 ties or races into which each species will be found divisible by cultural 

 selection. 



It has been found possible with many plants to increase the average 

 percentage of starch, sugar, or oil through the planting of selected 

 seed or cuttings, and there is every probabilit}^ that the same will be 



« Collins, Report on Caoutchouc, pp. 14 and 15. 



