Nov. 2s, 1918 , Biology of Fruit-Fly Parasites in Hawaii 457 



into the larva, no parasite developed and the fly matured and carried 

 in its body the encysted eggs of T. giffardianus . In those cases in 

 which both parasites oviposited, the larvse of O. fletcheri always died while 

 very young, and all of the larvse of T. giffardianus survived. One case 

 out of 201 occurred, however, wherein a female of O. fletcheri and 5 adults 

 of T. giffardianus emerged from the same puparium. During September, 

 October, and November, 1917, 3,092 adults of T. giffardianus were reared 

 from 201 melon-fly puparia by the foregoing method of double para- 

 sitism. After all of the adults of T. giffardianus had emerged, micro- 

 scopic examinations were made of the contents of each of the 201 empty 

 puparia. In ever)^ case the dead laK-a of O. fletcheri, usually one to a 

 puparium, was found. This was proof absolute that every host indi- 

 vidual had been stung not only by T. giffardianus but also by O. fletcheri. 

 In conjunction with this, 429 melon-fly larvae were stung by T. giffar- 

 dianus alone. These larvae all matured to flies, and each of the flies 

 when dissected was found to contain clusters of encysted dead eggs or 

 larvae of T. giffardianus in various parts of the body. 



As the duration of the egg of T. giffardianus is shorter than is that of O. 

 fletcheri, cases have been observed in which larvae of the former were 

 developing normally before the &gg of the latter had hatched. This 

 indicates that the immunization of the melon-fly larv^a against parasitic 

 development is destroyed at the time it receives the Q:gg of O. fletcheri. 



This weakened resistance of the melon-fly larva to parasitic develop- 

 ment is caused most probably by toxic substances injected into the larva 

 by the female of O. fletcheri during the deposition of the &gg. The reduced 

 resistance of the larva is not caused by any mechanical injury or bacterial 

 infection attending the insertion of the ovipositor. Larvae have been 

 pricked with fine cactus spines or heavily stung by Diachasma tryoni and 

 then exposed to attack by T. giffardianus. In such cases the eggs of 

 T. giffardianus invariably became encysted as usual and died. The eggs 

 even became encysted in larvae that had been almost killed by running 

 a cactus spine through the body. 



Unusual interest has attended the results of exposing melon-fly larvae 

 to the attack of O. fletcheri, followed by the exposure of the infested 

 melon-fly larvae to fruit-fly parasites other than T. giffardianus. In all 

 such cases the fruit-fly parasite eggs failed to develop, becoming as 

 tightly encysted as if the O. fletcheri egg had not been present. In other 

 words, no fruit-fly parasites will develop in melon-fly larvae under any 

 conditions, except in the case of the eulophid T. giffardianus, and this 

 species will do so only when combined with an egg or larv^a of the 

 natural melon-fly parasite, O. fletcheri. 



O. fletcheri will develop normally in fruit-fly lar\'ae. Laboratory 

 experiments have proved this positively and it has been reared from 

 fruit-fly puparia secured in the field in Hawaii on three separate occasions. 



