1919.] ROOT DISEASE OF CACAO IN TRINIDAD. 199 



In the Lesser Antilles, Rosellinia diseases occur in Guadeloupe, 

 Dominica, Martinique, St. Lucia, St. Vincent, and Grenada ; they are 

 unknown in the remaining islands of the group, which have drier 

 climates. 



The range of hosts is an exceedingly wide one, embracing practically 

 all the important cultivated and semi-cultivated plants, but the cultiva- 

 tions most affected are those of cacao (in all the islands), coffee (in 

 Guadeloupe and Martinique), limes (on new clearings in Dominicaj, and 

 arrowroot (in the interior districts of St. Vincent). 



The disease on cacao is usually caused by the species Rosellinia 

 Pepo, and in most cases at the present time, the fields being well 

 established, is communicated from the roots of dead or dying shade 

 trees, especially breadfruit, avocado and pois-doux. Another species, as 

 yet unidentified, is believed to attack cacao in certain localities. 



The disease on limes and coffee is caused by R. Pepo or R. bunodes, 

 indifferently. 



The first cases in new clearings are usually associated with forest 

 stumps left to decay, especially those of certain special trees. Subse- 

 quent cases arise from the spread of the fungus from tree to tree along the 

 roots, or by infection from surface soil rich in decaying vegetable matter, 

 which readily becomes infected in damp and shaded situations. 



An infected tree may be killed gradually by the progressive invest- 

 ment of the roots, or rather quickly by the destruction of the bark 

 around the collar. The fungus penetrates both bark and wood. Conidia 

 are produced with great readiness wherever the mycelium emerges into 

 the open : perithecia occur later, and especially in the case of R. Pepo 

 their formation may be long delayed. 



Much can be done to prevent outbreaks of the disease by measures 

 directed towards exposing to wind and sun the soil, the bases of the 

 trees, and anj' logs or other dead material Ij^ng about. Cases which 

 arise may be restricted by these measures and by the provision of a 

 close system of trenches to prevent root contact. 



Treatment by excision and exposure is successful in the early stages 

 of infection, but in practice these are hardly ever detected. Periodical 

 surveys would save many trees. 



Infected trees should be flamed and then promptly removed, all roots 

 dug out and burnt, the soil limed, and the situation exposed as much as 

 possible. All adjacent trees should be isolated from each other 

 by trenches. 



