19 



or the other wing lobe without variation. In Enkochckia, the 

 sacs are abdominal, variously placed, often several on one host. 



ABNORMAL LENGTH OF TIME SPENT IN THE 

 COCOON. 



Considering the whole number of species of Dryinidae known 

 to us, the average length of time spent from the period of spin- 

 ning the cocoon to the emergence of the mature insect, would 

 be from two to live weeks. Consequently as the Qgg state and 

 the larval stages are known to be of short duration, many gen- 

 erations can be produced in a year. There comes a time how- 

 ever when, owing to the absence of the host in a suitable stage 

 of development for the parasite that attacks it, or for other 

 reasons, the Dryinid itself has to assume a lengthy resting 

 period. So far as our observations go, this resting stage in the 

 case of the parasite always takes place when the larva has spun 

 its cocoon and before pupation. No doubt in countries witii 

 cold winters many pass the winter months in this stage; and in 

 warm countries this quiescent condition is liable to be assumed 

 at any season. During the summer months in Hawaii, occa- 

 sional larvae of Echtlirodclphax assume this torpidity in the 

 cocoon, remaining in this condition for months, though per- 

 fectly healthy; and in some of the winter months the number 

 that remain as larvae, when others (collected at the same time 

 and treated in the same manner) emerge after the usual period, 

 n,!ay amount to not less than 25 per cent. There is reason to sup- 

 pose that not only a lowering of temperature, but that dryness 

 even with increased temperature, may in some cases cause this 

 torpidity. About 2,000 cocoons of Dry inns, sent to Hawaii from 

 North America by Mr. Koebele, in the early part of November, 

 produced in Honolulu two males a few weeks after arrival, and 

 one male a month after these. The rest remained as torpid and 

 shrunken larvae in the cocoons for five months after they had 

 been collected. They were then taken to a cooler and damper 

 locality, and all that were not killed by hyperparasites produced 

 mature insects of both sexes in a short time. A number of curi- 

 ous cases were noticed in Queensland. For example, a larva of 

 a Paradryinus that spun up on Oct. ist, was still in the larval 

 state, and, as evidenced by occasional movement, alive on Dec. 

 1 2th, though much shrunken. It was then killed and preserved. 

 Larvae of this same Paradryimis that formed their cocoons in 



