\ 



1919. \ ROOT DISEASE OF CACAO IN TRINIDAD. 181 



The conidial fructifications are of the Graphium type figured by 

 various authors in connection with Rosellinia necatrix and other 

 described species. They occur previous to the development of 

 perithecia and are borne on the black surface mycelium which develops 

 wherever the fungus reaches the open under damp conditions. Each has 

 the form of a black bristle-like stalk 2 to 3 mm. long, built up of perpen- 

 dicular hyphae which branch out freely at the top into a tuft, which is 

 white or whitish to the naked eye from the conidia which cover it. The 

 conidia are borne laterally towards the terminations of these branches ; 

 the cells which bear them have a somewhat zig-zag or corkscrew 

 appearance from their tendency to bend away from the point of attach- 

 ment of a conidium. The conidia are rounded or oval, one-celled, 

 about 5 microns in length. 



The conidial fructifications are developed in very great abundance 

 (Fig. 2) and have been seen on dead leaves, twigs, and old lime skins 

 lying under close shelter beneath infested trees. 



The most striking characteristic of this species is the production of 

 fans or stars of white mycelium in the region of the cambium (between 

 bark and wood) of the roots (Fig. 3). The presence of these distinguishes 

 it at once from B. hunodes. A somewhat similar appearance may be 

 produced by the unidentified species (C), but in the examples of the latter 

 fungus seen, the growth has been very much less vigorous, 



On the roots the mycelium forms an irregular coating over the 

 surface. In its early stages it is smoky grey in colour, but soon becomes 

 black. It is gathered at first into rather loose branching strands with 

 spreading hyphae between them. Later the whole is combined into a 

 layer which is more or less woolly on the surface, and tending to be 

 carbonaceous below. It forms dense layers and pockets in the outer 

 bark, and from these whitish strands more or less vertical to the surface 

 everywhere penetrate the cortex. On reaching the surface of the wood 

 these repeatedly branch and spread in all directions over it, so that 

 when ' bark ' and wood are separated, a conspicuous white pattern of 

 branching lines, stars and fans is seen on the surface of the wood and 

 repeated on the inner surface of the ' bark.' From this layer strands 

 penetrate the wood in radial lines along the medullary rays, and send 

 out hyphae which invade and fill the large vascular elements. In this 

 species the mycelium in the wood, owing to its lack of colour, is not 

 apparent to the naked eye unless a cut surface is exposed for a day or 

 two, when the hyphae grow out and turn black. In long-infested wood, 

 thin plates, seen as black lines in a section, mark off certain areas, anrl 

 are believed to belong to this fungus. Such plates, however, are very 

 common in dead wood, and so have little diagnostic value. 



The external mycelium is most fully developed and conspicuous 

 when the fungus reaches the base of the stem and appears above the 

 surface of the soil. It is then seen in broad spreading fans or an 

 advancing sheet, which from the beginning or at an early stage encirclfS 

 the stem (Fig. 4). The margin while actively advancing is light grey 

 in colour for a width of about half an inch, behind which the colour 

 shades off to brown or black with a greenish tint. On lime bark the 

 sheet is smooth and glossy rather like wet fur. 



