i6 



latex in the tree. These improvements take, however, some time, 

 and the rubber planter of to-day cannot immediately benefit by the 

 gradual advances made in this connection. To him it is more 

 important to take advantage of methods of selection vv^hich can be 

 put into practice on his plantation already started. On such a 

 place it is of the greatest consequence that all unproductive, sickly, 

 or otherwise unsatisfactory trees be destroyed. It does not pay to 

 cultivate trees which do not give a payable quantity of latex, and 

 it is a waste of land to keep the ground occupied by such trees, 

 "which should be immediately replaced. 



In regard to the rapidity of growth and tlje size of the seedling 

 and the subsequent tree, it has long been conclusively proved that 

 the heavier seeds are far superior to the light ones. Experiments 

 conducted with a view of determining whether the size of the seeds 

 have any effect on the vigour of the plant, have shown that plants 

 grown from the heaviest seeds attain a greater size even if they do 

 not always germinate as rapidly as the smaller. It has further 

 been demonstrated that plants grown from thg heaviest seeds have 

 a greater power of resistance to drought. 



To insure a good stand and a greater yield, none but the largest 

 and heaviest seeds should be selected, and of the seedlings the 

 most rapid grower, with the healthiest and most vigourous 

 appearance. 



Experiments in Improving Castilla. 



So far, very few experiments have been made for the purpose 

 of improving the Castilla rubber tree, or in order to ascertain the 

 growth of the seedling under different circumstances. A series of 

 such experiments were started in April and May of 1906, at 

 La Zacualpa Botanical Station in Mexico. The results of these 

 initial experiments are not yet available for publication, but I have 

 found that so far a marked improvement is noticeable in seedlings 

 placed under favourable conditions and subjected to rational 

 and systematic treatment. 



Habits of Castilla. 



When we study the Castilla in its native conditions, in the 

 natural surroundings in which the tree has been able to exist and 

 flourish in competition with other trees, we learn many a useful 

 lesson as to its requirements under cultivation. It is by no means 

 essential that a cultivated plant should always have to be grown 

 under conditions identical to those of its original habitat, but we 

 can always derive pointers from a close observation of nature. 



One of the first things we notice when observing Castilla in 

 the wild state is that it prefers small openings in the forest and that 

 it never selects very heavy shade. In this regard, it is similar to 

 its relative the Guarumbo tree which is always found in localities 

 where the primeval forest has been cleared at some time or another. 

 The Guarumbo, or trumpet tree {Cecropia), is very common through 

 Southern Mexico and Central America. It is often called the false 

 rubber tree, because to the unitiated newcomer it resembles at a 

 distance Castilla. As soon as a clearing is made Cecropia will gain 



