19 



emerged was its own, or that of Haplogonatopus as it was at 

 first supposed to be. To determine this, more cocoons were 

 collected to ascertain whether more of this insect might be 

 bred. From 25 cocoons collected January 28, 5 Ceraphronids 

 emerged between February 1st and February 10th; then ex- 

 amination of remaining cocoons revealed 4 more of them con- 

 taining each a pupa which on rearing proved to be Ceraphronids. 

 This made 9 of the 25, or 36%. One pupa was with the re- 

 mains of a pupa of Haplogonatopiis, tending to prove that the 

 former is a parasite of the latter. 



Further observations were immediately undertaken in order 

 to establish proof in regard to this. Many cocoons of Ha,p- 

 logonatopus were collected at different times and examined. 

 The number which contained either larvae of Ceraphronid 

 feeding upon larvae of Haplogonatopus, pupae of Ceraphronids, 

 or from which adult Ceraphronids had already emerged, varied 

 from 24%) to 68%. In a few instances 2 pupae of Ceraphronid 

 were fomid in one Huplogonaiopus cocoon. In one instance 

 two larvae were found feeding externally on the larva of 

 Haplogonatopus within its cocoon, so it was made evident that 

 the Ceraphronid attacks its host within the cocoon. Fresh 

 cocoons of Haplogonatopus were obtained and a few Ceraphro- 

 nids admitted to them. They were observed apparently ovipo- 

 siting in these cocoons very soon, but the oviposition was not 

 actually seen. The female would traverse the cocoons several 

 times from end to end, all the time vibrating the antennae 

 rapidly and touching all parts of the cocoon, apparently to 

 determine whether the cocoon was an empty one or not, or 

 whether its contents were in the right condition for it to oviposit 

 in. After a few minutes she came to rest in a position which 

 would indicate that oviposition was taking place, even though 

 it could not be actually observed. This position was retained 

 for a minute or two. 



For further proof of the habits of this Ceraphronid, quite 

 a number of leaf-hoppers which were already attacked by the 

 larva of Haplogonatopus vitiensis, were collected from the 

 cane field, and freshly bred-out Ceraphronids placed with them 

 in breeding cage. None were observed to attack the leaf-hop- 

 pers themselves, nor the Haplog omit opus larvae preying on the 

 leaf-hoppers. The Haplogonatopus larvae were nearly full- 



