THE INSECT AS A PARENT 2C5 



which eventually eat their way out and spin their small, 

 yellow cocoons in a mass above the shrivelled remains of 

 their victim. The members of the large families Chal- 

 cididce and Proctotrypidce are less beholden to caterpillars 

 than the true ichneumons. Many of them oviposit in 

 pupae, many in the eggs of other insects and spiders, while 

 some patronise aphides. The extreme minuteness of 

 many of these insects may be judged from the fact that 

 in some instances half-a-dozen or more parasite grubs 

 find sufficient nourishment for their development in the 

 contents of a single butterfly's egg. 



Like the ichneumons, the Tachinid flies select cater- 

 pillars as hosts for their larva?. Their eggs are never laid 

 within the creature's tissues, however, but are fixed to its 

 skin by means of a gummy secretion. The minute larva 

 is left to penetrate the egg-shell on its underside, and 

 make its way into the victim's body. Moreover, the 

 instinctive shrewdness of the ichneumon far surpasses 

 that of her competitor the Tachinid. The latter frequently 

 glues her eggs to the back of a caterpillar that is about to 

 change its skin ; and when this happens the caterpillar 

 simply discards its old coat and the eggs along with it, 

 thus escaping destruction. Again, a fly will often attach 

 three or four times too many eggs to a caterpillar, with 

 the result that most of her grubs perish from want of food, 

 while the remainder are half-starved and produced dwarfed 

 imagines. Ichneumons, on the other hand, scarcely ever 

 make mistakes. They seem to know by a touch of the 

 antennae whether a caterpillar has been " stung " by some 

 earlier visitor, and if this proves to be the case, they rarely 

 insert their own eggs ; nor do they fall into the error of 

 overburdening a host. These comparisons suggest that 

 Tachinid flies are probably new to the business in which 

 they engage. We may suppose that, like many of their 



