ROOT DISEASES 131 



R. Pepo has formed a crust on the bark and is fruiting, some 

 blackening may extend to the mycelium under the bark at 

 that point. The examination of a root will remove any un- 

 certainty. In herbaceous plants the distinction between the 

 yellowish white strands of Pepo and the black ones of bimode'; is 

 usually quite clear. 



The disease on arrowroot is a special case, brought into promin- 

 ence by the estate cultivation of that plant in St. Vincent, of the 

 general effects of Rosellinia spp. on herbaceous plants with 

 succulent rootstocks. It is further described under the diseases 

 of that crop. 



Rosellinia {Paraguay ensis, Starb. ?) 



A species which agrees fairly well with the description of 

 R. paragiiayensis occurs on cacao trees in Grenada and St. 

 Vincent. The perithecia have also been found by the writer 

 on the wood of an Erythrina on a cacao estate near Soufriere, St. 

 Lucia. 



During a visit to Grenada in 1915 several groups of trees were 

 met with which were dying out, although growing in appar- 

 ently good soil and in favourable situations. The appearances 

 were such as to suggest root disease as the cause. The district 

 in each case was in the drier lowlands where the disease due to 

 Rosellinia Pepo seldom if ever occurs, and some of the character- 

 istic features of that disease were absent. The trees were slowly 

 dying back from the top, and putting out new suckers from 

 below, which in turn failed until the tree was completely dead. 

 In various stages of this process examined, the collar and upper 

 roots bore no sign of disease, but on the lower roots a mycelium 

 was found which formed a white radiating pattern between bark 

 and wood closely resembling that of Rosellinia Pepo but much 

 more scantily developed. These differences might have been 

 put down to the effects of drier conditions had not the finding 

 on dead cacao trees in two of these localities of the perithecia of a 

 distinct species of Rosellinia, associated with a similar scanty 

 white mycelium, brought the existence of a separate, though 

 closely allied, disease into question. 



In one instance of a diseased group, the only one on an other- 

 wise healthy estate, the trouble seemed to have had its origin 

 some sLx years previously in the felling of two examples of the 

 tree known in Grenada as tendre acailloux {Pithecolobium Berteria- 

 num, Benth.), in the others no such source was apparent. In 

 another instance bananas planted as cover where trees had 

 died in this way never reached the bearing stage, and examples 

 were seen of banana plants in various stages of a root disease 

 which may have been due to Rosellinia, but which could not 

 definitely be recognised as such. 



Some confirmation of the view that this species of Rosellinia 

 attacks cacao was obtained on an estate near Georgetown, St. 



