300 Annals Entomological Society of America [Vol. X, 



Before being fertilized the beetles would lay only a few scattered 

 eggs but in a day or so thereafter they would lay plentifully 

 and in good patches. Fertilization seemed to last several 

 weeks, but not for the season. One female observed was 

 found to be laying infertile eggs 35 days after being isolated 

 from a male. The spermatozoa of the later male seem always 

 to take precedence over all former, so that the eggs which 

 have not been already fertilized produce the characters of the 

 last male. The earliest egg patch was obtained April 1st and 

 the earliest beetles emerged May 1st. 



A few feeding records were taken on both bipnnctata- 

 annectans and annectans larvae. These were taken in very 

 warm weather, the thermometer being 87° to 93° each day. 

 The larvae, accordingly ate their niaximum and finished their 

 life cycle in the minimum time. In colder weather they ate 

 much less per day and the period of development was according 

 prolonged. These experiments were conducted with the greatest 

 care. The larvae were put into separate cages and the lice 

 which were given for feed were counted as carefully as possible. 

 Young of MyzHS circumflexiis were used for the first feed in each 

 instance and after that Chaitophorus negundinis was used entirely. 

 A check tube was kept to ascertain the number of lice dying 

 naturally in a day, but it seemed to be of little account, as 

 practically no lice seemed to die except from some disease or 

 from capsid injury, and this turned the dead bodies brown, so 

 that they could be easily distinguished in the larval cages. 

 Some of the larvae had already filled up on the unhatched eggs 

 of their patch before isolation, which of course did not count in 

 the food record, also whenever there was any doubt as to the 

 number of lice eaten the smaller alternative was taken. The 

 young lice that may have been born after being put in the cage 

 were not regarded, as they would not increase the bulk mater- 

 ially. The only difference they could make would be to add 

 to the number left over and subtracted, which would reduce the 

 number in the record, instead of exaggerating it. In these 

 ways every precaution was taken against getting too large a 

 count. The records are as follows: 



