ASCERTAINING THE PARASITES OP HOST-INSECTS. 79 



20 per cent., the emergence continuing during a considerably longer time, 10-12 

 days. In both these respects the curve of Aprostocetus strobilanae agrees entirely 

 with that of Torymus azureus, and the diagrams fig. 3 a and 6 show how closely one 

 curve follows the other. From this I draw the conclusion that Aprostocetus strobi- 

 lanae is the parasite of Torymus azureus. 



The above conclusions are of course based not only on the diagrams now published, 

 but on the great material brought together during the investigations of the spruce- 

 cone insects, which will be published later. 



The method was invented in order to enable me to ascertain the relation of the 

 cone insects to one another, and its usefulness seems to be proved by the fact that 

 the conclusion drawn from it regarding Perrisia and Platygaster by examination of 

 the cones was found to be true. 



If the material could have been kept under nonnal conditions, the curves would 

 evidently have been more separated from one another and the grouping of the 

 parasites round their respective hosts, as a consequence, still more plainly discernible. 



Very likely this method may also be applied with success in the investigations 

 of other complicated biocoenoses, when direct observations regarding the relation- 

 ship of the insects to each other are difficult to make, as for instance in large galls, 

 which are often inhabited by many different species, and tree-trunks. It is therefore 

 desirable that other entomologists should give the method a trial. 



