32 PESTS OF COFFEE 



when leaving the nest the ants would take the bugs also 

 along with the eggs and pupae. But while all the latter 

 were brought to the surface no bugs were brought up al- 

 though there were nearly twenty in the nest. In another 

 experiment the pot was flooded with water but with the 

 same result. 



These experiments give us valuable information as to 

 the relations between the ants and the scale insects. 

 There can be no doubt that those species which feed on 

 the honey dew have a large share in the distribution of 

 the pest. Considering how far the ants travel in search 

 of food and how soon they detect the presence of the 

 scales, their nests must be regarded as distributing centres. 

 This is probably the case even with species which make 

 their nests in the ground although, as already noticed, we 

 have not observed bugs being carried away from such nests. 

 The nests also appear to be convenient shelters from 

 adverse conditions of weather. Nests of Cremastog aster 

 in coffee estates were found crowded with green bugs in 

 September when elsewhere they could be detected only 

 after careful search or not at all. In fact, the nests seem 

 to be centres from which infestation starts after the mon- 

 soon is over. 



To determine the precise difference in the degree of 

 infestation between estates with and without ants' nests, 

 the following experiment was tried. Soon after the mon- 

 soon, two blocks equally badly infested and of about four 

 acres each were selected. One had all the ants' nests re- 

 moved or, as in the case of ground ants, destroyed by 

 carbon bisulphide, and for a considerable area round it. 

 In the other they were left undisturbed. After seven 

 months they were inspected again and it was found that 

 in the block where there were no ants' nests only eight 

 trees showed the bug while in the check there were one 

 hundred and thirty-two infested. 



As has already been stated, it has been proved by 

 experiments conducted in Java that ants prevent the 

 parasites of the bugs from approaching them. The bugs 

 attended by a,nts show less percentage of parasitised bugs 

 than those unattended. It has further been suggested by 

 van der Goot that the tickling of the bugs with the 

 antennae in some way helps to make them more vigorous. 



In South India, Fletcher records having observed the 

 ant, Gamponotus compressus, Fabr., vigorously defending 



