DEVELOPMENT OF CYNIPID HYPERPARASITES 478 
the viscera of the aphid, and then secretes silk and weaves the 
cocoon as usual. The tissues retain their tone and colour, and 
irritation excites slight movement. On close examination, 
however, it can be seen that the body is somewhat contracted. 
At this time the Cynipid larva, its head orientated with 
that of the host, lies above the mesenteron of the latter, which 
it constricts into a dumb-bell form. By some means the further 
development of the Aphidius is arrested, and always at 
the same point, namely, after the weaving of the cocoon. 
The meconium is never evacuated, and metamorphosis, which 
normally takes place soon afterwards, never occurs. ‘The con- 
dition of the Braconid larva resembles in fact that of the prey 
that certain Hymenoptera store in their brood-cells. 
Two explanations of this phenomenon suggest themselves. 
Bither the female Charips at oviposition may inhibit the 
final changes of the host, possibly by injection of some secre- 
tion; or the Cynipid larva itself, during development, may 
affect the Aphidius by chemical or physical means. 
The evidence is not conclusively in favour of either view. 
In support of the first one particularly marked instance came 
under notice. 
A Charips female was observed to oviposit on June 26. 
The aphid was isolated, and four days later the Aphidius 
within began to spin silk. On July 4 the cocoon was opened 
in order better to follow the development of the hyperparasite, 
a plan that was adopted successfully in several instances. The 
A phidius remained without change until August 7, a period 
of five weeks. The meconium was not voided, but beyond some 
contraction the larva looked healthy. In replacing it in the 
tube after examination it fell from the brush, and must have 
received some injury, for next day a discoloured patch appeared 
at the hinder end of the body. The larva was dissected care- 
fully, but no hyperparasite could be found, and the organs 
showed little signs of histolysis. As oviposition had been 
observed, the facts suggest that some accident had prevented 
the development of the Cynipid larva, and this leads to the 
inference that the agent arresting the metamorphosis of the 
