156 FROGHOPPEll BI.IGHT OF SUGAR-CANE. 



never been recorded before October. It can only be cultivated with 

 difiBculty and does not lend itself to artificial distribution. On occasion 

 it has destroyed large nuirbers of froghoppers and infected specimens 

 are killed more i-apidly than by the green muscardine (p. 86). 



Hypek-parasites. Under this heading are included the animals 

 which destroy the enemies of the froghopper and so allow it to increase. 



The mongoose has been held responsible for the increase of the 

 froghoppei's by destroying their enemies. Comparisons of the food of 

 the mongoose in districts where it has been long abundant with districts 

 where it has only recently appeared, do not show any change of diet, 

 which would be expected if it had been responsible for the 

 extermination of birds, lizards, frogs and toads. Blight was known 

 before the introduction of the mongoose and the outbreaks of 

 recent years have not been correlated with the spread of this animal. 

 The conclusion reached is that the mongoose, although doubtless 

 undesirable, and possibly to a very limited extent contributing to recent 

 outbreaks of pests, is not the cause, nor is it even one of the most 

 important causes, of the outbreaks of froghopper blight in 

 Trinidad (p. 87). 



The mongoose eats a number of rats and these can climb and have 

 been found destroying the nests of insect-eating birds. 



Birds, lizards, ants, spiders and other general insectivorous enemies 

 of the froghopper sometimes do harm by destroying its insect 

 enemies (p. 91). 



The spider and grasshopper enemies of the froghopper are destroyed 

 by fossorial wasps, while the former are also killed by an entomogenous 

 fungus {Gibellula) (p. 91). 



A worm believed to be parasitic, has been found in the pupa of a 

 Syrphid fly (p. 91). 



In Fig. 19 (p. 92) an attempt has been made to show the relation of 

 the froghopper to its parasites and hj'per-parasites. In the centre the 

 froghopper is shown and in the first ring round it all its natural enemies 

 grouped according as they feed on the egg, nymph or adult. In the 

 outermost ring are the hyper-parasites or enemies of the enemies, and 

 in the intermediate ring are shown other hosts or foods of the parasites 

 and hyper-parasites. The arrows indicate that the animal at the tail of 

 the arrows destroys the animal at the head. A careful study of this 

 diagram will show the extraordinary complexity of the relationships and 

 the great difficulty of deciding whether any particular animal is desirable 

 or undesirable, (p. 92). 



THE CANE ROOT FUNGI. 



Two distinct kinds of root fungi are found attacking the canes in 

 Trinidad. 



The Marasmius type mats together the lower sheaths of the cane 

 with a white dried-paste-like mycelium (PL VII) and produces spores on 

 small mushrooms (PI. VIII). The Odontia-Himantia fungus mats the 

 lower sheaths of the cane with a feathery mycelium (PL IX) and produces 

 epores on a rough granular surface growth on the outside of the sheath 

 (PL X) (p. 93). 



Both these fungi can exist in a healthy field of canes living only on 

 decaying vegetable matter and doing no harm. There is hardly a field 

 in Trinidad where traces of one or the other cannot be found after a 

 close search. 



Under certain little understood conditions, particularly when the 

 canes are suffering from other causes, these fungi become parasitic and 



