never develop but remain in the form of larger or smaller scales. 

 At the apex of every fruit is a small hollow in which the dried up 

 remnants of the style can be seen. 



The Best Type of a Castilla Tree. 



We have yet to ascertain the cause of the difference in yield of 

 individual trees, or at least, we must find out what trees are the 

 best producers and the most rapid growers, before we can enter 

 upon the problem of determining or developing the best type for a 

 rubber tree. On general principles we should require a tall, 

 straight trunk, with a dense crown at the top of the tree. But the 

 natives hold that a tree which is not too tall and which has a taper- 

 ing trunk, gives a larger yield -and better rubber than a taller tree 

 of the same age and with the same girth at the base. I have been 

 able to verify this by actual experience, but I cannot give any 

 explanation of this fact. 



The object is to develop a trunk with as large an area as pos- 

 sible for tapping. Branches seldom attain a tapable size and a 

 rubber stand with clean stems, without intertwining branches or 

 underbrush, is easier to work in than in a tangled mass of trees 

 growing without order and care. 



We have many instances of large trees with tapable branches, 

 or trees which have branched from the base, being regularly tapped 

 and producing a quantity of latex. It may seem an advantage 

 to have several stems to tap instead of only one, where a single 

 shoot or a single trunk has developed, but there can be no question 

 as to the fact that where several branches are allowed to develop the 

 growth of the main trunk is greatly retarded. 



It is therefore necessary to have a tree straight and clean 

 boled, in a condition to make the best of the period of maximum 

 growth, the time of which has not yet been ascertained. We 

 know for a fact that the best rubber producer has a thick, compact 

 crown. The conical form of crown is also to be recommended as 

 it naturally receives more light than a'flat crown. After the Castilla 

 growing in the wild state has obtained its maximum height the 

 crown always becomes flat and rather ovate in shape, while in 

 youth, when the tree is growing vigorously under normal condi- 

 tions, it ha? a sharply conical crown. Every kind of tree has a 

 maximum height to which it is able to pump water, and when this 

 height has been reached the growth of the tree ceases because the 

 crown cannot be supplied with sufficient water. The normal 

 rubber tree should not, therefore, be very tall as in the best situa- 

 tions the wild tree reaches a height of about sixty feet, and the 

 over-mature trees always have a very spreading crown. Some- 

 times dry topped young trees are observed. This is due to un- 

 suitable conditions in some respect or another, and we recognize 

 this as a disease, called by foresters the " staghorn disease." 



The leaves should be large, with a fresh green colour, the bark 

 thin and smooth. Some trees have leaves with stiff, bristly hairs, 

 and I have found on some plantations almost every tree covered with 

 these stiff hairs, sometimes resembling prickles. In cases where 



