a foothold and as it is of very rapid growth, it soon grows into a 

 small tree. I have noticed on some plantations that the managers 

 studiously avoid cutting down the Guarumbo trees wherever they 

 grow among the rubber. I was informed that this was done on 

 purpose as the Guarumbo tree resembled the rubber tree so much 

 that it helped to carry out the impression of an even and good 

 stand of rubber trees. Inspecting shareholders did not notice the 

 difference. I am ready to believe this, as there appears no other 

 reason why the Guarumbo tree should be left standing. It is not 

 good as a shade tree and as it consumes a great amount of water 

 and plant food, it is decidedly detrimental in a rubber stand. 



In the natural succession of the forest trees both Cecropia and 

 Castilla are secondary elements. The seed of Castilla is so thin 

 shelled and perishable that it needs a moist place in which to 

 germinate as it would otherwise be destroyed by the heat of the sun. 

 We therefore always find the young seedlings growing close to 

 other trees which give them sufficient protection. 



Castilla depends in many other respects upon its neighbours 

 for safety. It is a very brittle tree, easily broken by the wind 

 and therefore needs a wind-break. We always find Castillo* in the 

 pole stage standing close to other trees, which indicates that they 

 have been able to survive only by reason of the protection afforded 

 by the other trees. It may be inferred that if this is true we would 

 never find a Castilla standing separated from any neighbours. I 

 think that in every case where an old Castilla tree is found single 

 it has either been planted by man and protected, or, if a wild tree 

 the surrounding trees have been destroyed in some way or another. 

 It k at least on very rare occasions that Castilla is able to live 

 through the sapling and pole stages without protection against wind. 



Another way in which Castilla gets protection from neighbour- 

 ing trees is that these give shelter to the ground, retain moisture, 

 and prevent the soil from cracking. Castilla is very sensitive to 

 these influences, as its roots do not develop properly in hard-baked 

 ground. 



Castilla does not, on the other hand, develop well in shade. 

 It grows very slender, with a weak trunk and an undeveloped 

 crown. The tree needs plenty of light for its foliage and it is only 

 where the Castilla tree has room enough to spread its branches 

 and expose its foliage to the rays of the sun that it can successfully 

 maintain its position in the struggle for space and light. 



In districts were a distinct dry season prevails, Castilla is 

 pronouncedly deciduous and drops its leavesat thatperiod of the year, 

 while in a humid region the shedding of leaves goes on all the 

 year round. On the Pacific side of Sierra Madre in Mexico, the 

 dry season lasts from January to May, and the rubber tree begins 

 to shed its leaves with the advent of this season, and towards its 

 close the trees are almost destitute of foliage. 



It is generally stated that Castilla does not flower and set 

 fruit before it has reached an age of about five years. Another 

 statement is that flowering commences when the tree begins to 

 develop permanent branches. My observations show that if 

 Castilla is grown on good soil in a suitable climate, and if the 



