ROOT DISEASES 145 



to include some line of infection that has run off in advance of 

 the general spread. A circle leads nowhere, whereas a system 

 of squares may be added to at any point and be carried in any 

 direction, and is capable of any subdivision. It has moreover 

 the great advantage of linking up with an existing or prospective 

 drainage system. 



As regards the form of the trench itself, there is but one 

 essential so far as root disease is concerned : that it should be 

 deep enough to cut through all the roots passing across its 

 situation. When digging near an infested tree, the earth re- 

 moved should be thrown inward as a precaution against the 

 scattering of possibly diseased material among the healthy trees. 

 This refers more particularly to secondary trenches ; the outer 

 trench should be put far enough away to avoid, in general, 

 the chance of finding such material. The earth should be dis- 

 tributed, not banked at the edge of the trench, unless put there 

 in exceptional circumstances with the express purpose of avoiding 

 wash. Under no circumstances, however, should the collar of 

 the tree be earthed up. Roots passing into the healthy area 

 should be followed up and removed, so far as this can be done 

 without much injury to other roots. It is better for cut ends 

 of roots to be protected with a coat of tar. 



Survival of Supply Plants. 



On the thoroughness with which the clearing up is done de- 

 pends the chance of survival of an early supply plant. It has 

 been found in St. Lucia, where the results of several years of 

 experience of the treatment of this disease on cacao are now 

 available, that, where the work has been carried out under the 

 personal supervision of the planter, supplies put in a month 

 afterwards have remained healthy. The longer the delay the 

 greater the chance of survival. It must be remembered that a 

 supply may do well for a few months while its roots occupy the 

 site from which the stump was removed, and then become 

 infected from some outlying fragment of the old tree as its 

 roots spread wider. Such cases have been definitely traced. 



The Possibilities of Control. 



Rosellinia disease in orchard cultivations is by its nature 

 capable of the most serious consequences. It is cumulative in 

 its effects : each tree that contracts it infects as a rule not merely 

 one but several more, and each tree killed takes at least five or 

 six years to replace. In new clearings, encumbered with stumps 

 and logs, and in cacao fields with large shade trees, the disease is 

 difficult to deal with. 



It may be confidently claimed, however, that the application 

 of the principles of prevention and control set out in the preceding 

 pages will reduce the annual losses to a minimum that is not 

 likely to be serious, and one may further expect a gradual 

 diminution year by year of even the sporadic cases. 



L 



