DEVELOPMENT OF CYNIPID HYPERPARASITES 453 
generally, and to indicate the particular sub-genus or species 
where necessary. 
I would here express my sincere thanks to Professor J. Stanley 
Gardiner for giving me facilities to carry out the work in the 
Zoological Laboratory at Cambridge ; and my obligations to 
Professor J. J. Kieffer, who kindly determined the examples 
of Cynipidae submitted to him. 
MATERIAL. 
The material used was obtained in Cambridge in the 
summer of 1920. Charips (Allotria) has been reared 
from various Aphidiidae in different aphides, but through- 
out this work, Aphidius ervi, Hal., a parasite of 
Macrosiphum urticae, Kalt., was used, as the com- 
paratively large size of the cocoons rendered them con- 
venient for dissection. The parasite and its host were 
common and widely distributed round Cambridge in June 
and July. Moreover, the food plant of this aphid, the common 
nettle, usually grew in isolated patches along the roadside. 
This was an advantage, since the Aphidius, after parasitiza- 
tion by the Cynipid, is liable to secondary parasitization by 
certain ecto-parasitic Chalcids and Proctotrypids, which kill 
both the host and the first hyperparasite. Collections made 
from one spot showed that almost every Aphidius, whether 
attacked by a Cynipid or not, might bear one or more of these 
external parasites ; while collections made fifty yards away 
were free from secondary infestation, and contained Cynipid 
larvae in all stages of development. 
The rearing methods were the same as those used when 
studying Lygocerus (10). Camera lucida drawings and 
measurements were made from living specimens, mounted in 
salt solution or dilute glycerine. ‘The larva, and the host 
when necessary, were also studied in serial sections. 
Brotocican Notr on tHE Host. 
The development of the Braconid, Aphidius, within the 
aphid has been deseribed by Seurat (21), Timberlake (25), and 
others. 
Hh2 
