Nov. 25. 1918 Biology of Fruit-Fly Parasites in Hawaii 



421 



These pass through parts of three nights and two days as opposed to 

 three days and tv.'0 nights and would thus be subjected to somewhat 

 lower average temperatures. This will explain some of the misleading 

 variations shown in Table I. 



Table I. — Duration of the egg stage of Diachasma tryoni in Honolulu 



The fully developed egg is so swollen and the membrane so thin that 

 actual emergence of the larva from the egg is sudden, and rather an ex- 

 plosive process. Many eggs on hatching have been under observation. 

 The egg membrane is suddenly ruptured longitudinally, probably by 

 the movements of the larva within, and the larva floats out without ap- 

 parent effort into the semiliquid medium of the host surrounding it. 

 The egg may hatch while the host is still in the active, feeding, larval 

 stage or it may hatch after the puparium is formed and the complete 

 histolysis of larval tissue has taken place. No important histogenetic 

 action *t)ccurs in a puparium containing a parasite egg or larva. The 

 parasitized host larva feeds and develops to maturity even though heavily 

 superparasitized, leaves the fruit normally, and forms a perfect pupa- 

 rium in the usual manner. The complete histolysis of the larval tissues 

 within the puparium then follows, but here all fly development ceases. 

 Henceforth the content of the puparium is but a liquid mass containing 

 the broken-down bits of larval tissue and the rapidly developing para- 

 site larva. The death of the host then may be said properly to occur 

 at the cessation of histolysis in the newly formed puparium. From over 

 3,000 parasitized puparia opened during 191 6 and 1917 no single case 

 was ever noted in which a perfect or even partially formed fly pupa 

 occurred. 



LARVA 



The larva midergoes many interesting phases in the processes of 

 transformation. When first hatched, it is about 0.85 mm. long. This 

 is the most active period in the larval life. At this time it is so markedly 

 different from the later instars that it appears to simulate the larval 

 structure and habits of an ancestral type. It usually hatches while the 

 host is still in the larval stage. The parasitic larva then lies in a well- 



