578 BIOLOGY OF CHALCIDID/E HOWARD. 



These iustances will suffice to indicate the extreme rapidity of growth 

 of mauy of these parasitic larva). The question of number of annual 

 generations is, I believe, entirely one of appropriate food. Copulation 

 takes place immediately after the adults issue, the males usually appear- 

 ing a little in advance and awaiting with impatience the egress of the 

 females. Very soon after coition the females are ready to oviposit, and 

 in the case of polyphagic species or species which attack insects of 

 great abundance whose generations overlap there must be mauy so- 

 called "broods" in a siugle season. 



Where host-insects are not accessible, however, there can be no 

 doubt but that the impregnated female can live a long time, and hiber- 

 nation in this state is frequent. Another common method of hiberna- 

 tion is in the full-grown larva. Those species which issue from the 

 pupa) of Lepidoptera usually overwinter in this condition, transform- 

 ing to a short pupal stage in the spring. 



A curious fact, and one contradictory to the usual rapid development 

 of these insects, is given by Scudder (Butt. New Eng., p. 701), who sur- 

 mises that the pteromaline parasitic upon Euphydryas pliaetoyi possibly 

 requires two years to complete its transformations, since all of the 

 chrysalids of the butterfly which hang through the winter are parasi- 

 tized. The butterfly larvae it seems hibernate and transform to chrys- 

 alids in May and June, giving out the butterflies in June. When 

 parasitized, however, they hang all summer and through the following 

 winter, the parasites appearing on the wing the following June. While 

 it appears to me that the possibilities iu this interesting case are over- 

 stated in supposing that the eggs of the parasite are necessarily laid 

 upon the larva) of the butterfly iu late summer and fall, it is still re-' 

 markable, for, even on the supposition that the ^g^ is laid in the 

 chrysalis, the parasite must occupy a full year in development, always 

 providing the facts stated are strictly correct. 



The preceding remarks apply strictly to the parasites of external 

 feeders, for with parasites of endophytes the i)eriod of development is 

 undoubtedly longer. With gall parasites, for instance, I believe that 

 there is never more than one annual generation, for the galls them- 

 selves are of annual development and must be pierced at a certain 

 stage of their growth. In such cases, moreover, there may be a retar- 

 dation of development due to absence of natural moisture, as where 

 galls are kept dry indoors. In such cases Ashmead has shown (Proc. 

 Entom. Soc. Wash., I, 91) that cynii)ids may be retarded for two 

 years and then brought forth by the application of water to their 

 galls. The same laws will undoubtedlj' apply also to their parasites. 



It may also be worth recording here that Mr. A. Craw, of Los Angeles, 

 Cal., considers that Dilophogaster California mihi has but one annual 

 generation. This insect is a parasite of the Black Scalcof California 

 {Lecanium olca'), and destroys annually 75 per cent of these scales. Ac- 

 cording to Mi\ Craw the parasite deposits eggs\in the uuiture scales 



