1919.\ ROOT DISEASE OF CACAO IN TRINIDAD. 183 



Fungi are of course very abundant on the decaying material, but the 

 Rosellinias are far from being conspicuous among them. I have 

 once found in Dominica a large decayed log infested with the mycelium 

 of Rosellinia bunodes, as shown by the infection at the point of contact 

 of lime branches touching it, and infection of a shrub growing upon it, 

 and have found three other species on Erythrina logs in a St. Lucia 

 cacao cultivation. 



Nor have the Rosellinias been found so often as might have been 

 expected on forest stumps. In a fruiting condition they are in fact 

 rare in this situation. The most notable case T have seen was that of 

 a chataignier stump {Sloaiiea sp.) covered with perithecial and conidial 

 fructifications of R. Pepo, which from its position appeared to be the 

 centre of infection of a large group of diseased limes. The presence of 

 the mycelium on the buried roots of forest stumps is more frequently 

 observed. Every infested clearing shows examples of the association of 

 diseased trees with stumps, and sometimes these are very striking, as 

 when a group of five or six dead and dying trees is seen around a large 

 chataignier. The actual nature of the connexion is usually difl&cult to 

 demonstrate ; as a rule the cases when seen are too far advanced for 

 sure conclusions, and often they have to be passed over for want of 

 time or means for the uncovering of the roots. But from first to last 

 a good many cases have been investigated in which the roots of diseased 

 trees have been found in contact with roots of forest stumps bearing 

 the fungus. In some of tliese the evidence that the fungus passed from 

 the stump to the lime tree is quite definite. Two examples in which 

 the evidence includes the identity of the forest tree may be cited. 

 South (1913) reports having followed up the fungus from dead mahoe 

 cochon (Sterculia caribaea) to the roots of a living lime tree; the writer 

 examined a large stump, still bearing a few living suckers, of 

 chataignier grand-feuille {Sloanea Massoni) the roots of which were 

 badly infested and had recently communicated the fungus to the roots 

 of a lime tree in contact with them. In other cases a single root coming 

 from the direction of a stump has been found to be infested when all 

 others were sound. From the nature of the case, examples in an early 

 stage are only found by accident or a lucky shot, so that one is justified 

 in assuming that a fair proportion at least of infections near stumps 

 take place in this way. 



The fact remains that considering the area of the clearings, the 

 number of infested stumps does not appe ir to be large. Relatively few 

 trees are lost in the first three or four years after planting. This is in 

 part due to the time taken (a) by the fungus to develop on and about 

 the stumps, (b) by the roots of the planted trees to grow out into the 

 infested area, and (c) by the fungus to kill a tree after infection. But 

 observation shows that when the period of heavy losses ensues, most 

 of the trees are infected from one another, so that the characteristic 

 distribution of the disease is in scattered patches whose number is very 

 small compared with that of the stumps and logs in the clearing. 



There is strong reason to believe, and the idea is supported by 

 experience in other countries, that the number of forest trees whose 

 stumps readily serve as centres of distribution for the fungus is quite 



