— 31 — 



2. Dorsally right on metathorax, 



Ventrally in the médian line between 4. and 5. abd. segm. 



3. Dorsally right, between metathorax and I. abd. segment. 



4. On the side between metathorax and 1 . abd. segm. 



Ventrally right, between 3. and 4. abd. segment. 



5. Dorsally right, on 2. abd. segm. 



6. Ventrally left, on 2. abd. segment. 



7. Dorsally in the médian line between 2. and 3. abd. segm. 

 Ventrally right, between 5. and 6. abd. segm. 



8. Ventrally in the médian line between 2. and 3. abd. segm. 



9. Dorsally in the médian line, on 4. abd. segm. 

 10. Dorsally left, between 5. and 6. abd. segm. 



From this list it can be seen that the oviposition cannot take 

 place, the mcther wasp having put the egg in through the ante- 

 rior opening of the caddis-fly case, surely the caddis-fly larva 

 would défend itself, the head and prothorax can completely 

 cover the opening, and, moreover, the eggs are placed so far 

 backwards on the host larvae that the short ovipositor of Agrio- 

 typus could not reach so far. The mother wasp thus certainly 

 must penetrate the side wall of the case, the sting is strong 

 indicating that it is used in rather hard work, and the above 

 list shows that most of the eggs are placed on the dorsal side 

 which is of course the easiest to attack. 



The young larva (fig. 5 B) of which I hâve found only one 

 spécimen, lives entoparasitically ; it is rather hyaline, 13-seg- 

 mented, on the 10 anterior segments a dorsal row of bristles of 

 which the médian ones are the longes t. The 3 last segments 

 without bristles and the segments only indistinctly separated. 

 The head seen from above (fig. 5 C) shows two oblique lines 

 ending blindly posteriorly and corresponding to two internai 

 ridges. Posteriorly to them the antennae are seen as two little 

 warts. Seen from belcw (fig. 5 D) the head appears furnished 

 with two hooklike mandibles which exteriorly will articulate on 

 the fore end of a long chitinous ridge. The gut ends blindly, 

 no tracheae or spiracles can be observed. Length 1,4 mm. 



The mature larva already being found very early in summer 

 (my first date is 8/6), the first stage therefore only can be found 

 during a week or at most in the course of a fortnight, and this 

 is surley the reason why I did not succeed in finding more spéci- 

 mens of this stage. 



