39 



stem or branch. Often two patches unité into a single 

 one, or one first appears under the surface and joins 

 itself on to another, in the latter case tlie infection must 

 hâve spread from within to the outside. Even in badly 

 affected trees, spots which may penetrate to the wood, 

 are not always a deep claret colour, but often light red. 

 When thèse lightcoloured patches are exposed to the air 

 after cutting, they become darlc red. Where the wood is 

 also affected, it sometimes assumes a red, but generally a 

 Wackish brown colour which may penetrate, into the wood 

 for some centimètres; fig. 2 and 3 give an illustration of 

 the more common case, in which only a small border of 

 the wood is discoloured. In fig. 3 the progress of the 

 discolouration in the bark is clearly visible. This dark dis- 

 colouration of the wood is sometimes continued in nar- 

 row stripes far under the healthy bark. 



In cutting out pièces where the wood also is diseased, 

 one often flnds bark and wood qui te separated from each 

 other, even where there is no question of insects having 

 entered. Sometimes a gummy liquid has accumulated 

 between the two. 



How long the canker takes to kill a tree I cannot say 

 with certainty. It is probable that no more than a few 

 months is required, for in July many trees were found 

 dead, which partly at least had most likely only been 

 affected in the rainy season, but further observation on this 

 point is necessary. Sometimes a cankerspot can be traced 

 to hâve spread from a wound, but the roughness of the 

 bark often makes it impossible to ascertain this. In rare 

 cases there is a cankerspot at the footof dead „krulloten". ') 



1) Krulloten are hypertropliied twigs, due to Colletotrichuni hixi- 

 ficum. C. J. J. van Hall et A. W. Drost. Les balais de sorcière 

 du cacaoyer provoiiués par Colletotrichuni luxificum n. sp. Recueil 

 des Travaux Botaniques Néerlandais. Vol. IV. 1908. p. 243. 



