PROPAGATION BY CUTTINGS. 59 



of the tracts which represented an experiment in open planting had 

 each young- seedling covered with a hood made of leaves of the manaca 

 palm (Attalea). 



PROPAGATION OF CASTILLA FROM CUTTINGS. 



Under favorable conditions Castilla roots readily from cuttings. 

 Between Pueblo Nuevo and Huitzla, along the road from Tapachula 

 to La Zacualpa, there is a row of several trees planted as fence stakes, 

 but at present flourishing in an entirely normal manner. That trees 

 raised from cuttings are not stunted or otherwise abnormal is also 

 shown by many other instances in the neighborhood of Tapachula. 

 Indeed, this was the origin of the largest tree observed (PI. IX). It 

 measured 7 feet in circumference about 5 feet from the ground. 

 It was about 33 feet to the first limb and the total height was estimated 

 at 80 feet. The owners claim that it has yielded 6 pounds of rubber. 

 It had apparently not been tapped as persistently as trees at La 

 Zacualpa, perhaps because protected from "huleros" by standing close 

 to a house. Milk flowed freely from a knife cut, and appeared to be 

 rich in rubber. Natives claimed that an arroba (about -25 pounds) of 

 rubber could be obtained from such a tree by the methods used by 

 them in the forest. This tree was further notable in that it had 

 apparently grown up in the open and without near neighbors. 



The propagation of Castilla from cuttings has already been men- 

 tioned as the quickest method of bringing about the selective improve- 

 ment of Castilla. It is not known that any plantations have been 

 stocked with cuttings, presumably because in localities where they 

 could have been obtained seedlings could be had more readily and in 

 larger numbers. Indeed, wood suitable for cuttings does not exist 

 in very large quantities even in a well-grown tree, because the false 

 or temporary branches can not be used for this purpose. Like the 

 lateral branches of some of the coniferous trees, they are unable to 

 give rise to normal terminal buds, so that even if false branches could 

 be rooted they would not produce normal trees. The failure of such 

 experiments has been reported from British India. 



Planters resident in regions in which wild rubber trees exist could 

 try a useful experiment by planting in rows cuttings taken from 

 different wild trees known to be above or below the average in pro- 

 ductiveness. This would enable them to determine whether the dif- 

 ferences of yield were due to the external conditions or to causes 

 inherent in the trees themselves. The probability is that very appre- 

 ciable individual differences of productiveness will be found, and that 

 a distinct advantage can be secured by using cuttings of such trees 

 for propagating purposes. 



The fact that stakes 5 or 6 feet long and 2 or 3 inches through take 

 root and grow when simply set in the ground shows that propagation 



