286 Journal of Agricultural Research voi. xii, No. s 



and in its natural environment throughout the whole range of its existence and a 

 similar biological study of its parasites and natural enemies under such conditions. 



Some results of the recent work in Hawaii most strongly bear this out. 

 Sufficient evidence has been obtained to throw serious doubt upon the 

 assumption, often accepted, that the greater the number of species of 

 parasites associated with the host, the greater the chances for its 

 control. It is felt that the following data quite definitely indicate, at 

 least in some cases, that better results may be obtained by a method of 

 judicious selection of desirable species for introduction rather than by a 

 wholesale and indiscriminate procedure. If 90 per cent of all individuals 

 of an insect pest are destroyed by a single species of parasite, is it wise 

 to attempt further control by bringing in other species, until it is known 

 by positive and careful experimentation in the laboratory and field that 

 these new species will not interfere with or check the normal activities 

 of the first species ? 



In May, 191 3, Prof. F. Silvestri succeeded in bringing two species of 

 opiine parasites of the fruit fly into Hawaii. One, Opius humilis Sil- 

 vestri, he brought from South Africa, and the other, Diachasma tryoni 

 Cameron, was secured in Australia. Both were soon established in the 

 Kona coffee district of the island of Hawaii. By 191 5 it had become 

 clearly evident that O. humilis was often parasitizing from 60 to 90 per 

 cent of all of the fruit-fly larvae developing in the coffee cherries. D. 

 tryoni steadily but slowly increased and in time exhibited a capacity for 

 occasionally parasitizing 50 per cent or more of the host larvae. Here 

 it is obvious that overlapping or duplication in parasitism was occurring. 

 Clearly some fly larvae were being stung by both species of parasites and 

 frequently to a very considerable extent. 



CANNIBALISM AMONG THE PARASITES 



Early in 1916, Dr. E. A. Back, of the Bureau of Entomology, while 

 examining the contents of some parasitized fruit-fly material from the 

 field, observed under the microscope a larva of the parasite Diachasma 

 tryoni attacking one of the parasite Opius humilis. A suspicion of dis- 

 advantageous consequences arising from complications attending the 

 interactions of these parasites led Mr. C. L. Marlatt to assign to the 

 writers an investigation of this subject. 



Careful microscopical examinations of large numbers of fruit-fly larvae 

 and pupae, collected from localities where both species of parasites were 

 known to be well established and actively working together, soon revealed 

 one striking fact. In the majority of cases where fruit-fly larvae had been 

 parasitized by both Diachasma tryoni and Opius humilis the latter was 

 killed and the former developed to maturity. O. humilis is killed purely 

 by wounds and lacerations inflicted upon it by the long, curved, sickle- 

 like mandibles of the newly hatched larva of D. tryoni. This larva 



