INSECT PESTS OF PINE AND SPRUCE CONES, CXLIII 
Aprostocetus strobilanae occurs all over the country (map fig. 40). Table 
IV shows the occurrence of Torvmus azureus and Aprostocetus strobilanae in 
spruce-tree cones, 200 from each locality, collected during the winter 1915- 
1916. Only in two localities was Aprostocetus present although Zorymus did 
not occur. But in both localities Aprostocelus was very rare, only 1 and 6 
specimens respectively being found; and it is therefore possible that in both 
cases all Zorymus were parasitized by ÅAprostocetus. 
The diagram, fig. 41, shows the distribution of the localities according to 
the percentage of Zorymus parasitized by Aprostocetus. A comparison between 
this diagram and that of Perrisia strobi and Platvgaster contorticornis (fig. 29) 
shows that ÅAprostocelus is a very effective parasite, which in the material in- 
vestigated surpassed Platvgaster. While in 63 & of the localities the latter 
only had killed up to 20 2 of Perrisia, the corresponding figure for Aprosto- 
celus is 30 Z of the localities. The curves of both species go down at 
50-60 ?; infection; but while the highest percentage of the localities in Platy- 
gaster is to be found at o 4 infection, the corresponding figure of Aprosto- 
celus 1S at 30-40 & infection. 
The curve of Aprostocetus, however, is at the higher percentage of infection 
very peculiar, rising from 71-80 2 so that not less than 12 & of the loca- 
lities show an infection of 91-100 2. 
It is, in itself, not inconceivable that the degree of infection may be so 
high in so many localities, but, on examining more closely the entire data 
for these localities, which are recorded in table V, another explanation pre- 
sents itself. | 
We notice that it was precisely in these localities that Perrisia strobi Was 
very numerous, while at the same time Platvgaster either was very scarce or 
did not occur at all. This coincidence seems to signify that some relation 
exists between these features. 
The larvae of Perrisia and Torvmus both live in the seeds of the spruce- 
tree and compete accordingly in their food-habits. The data of breeding 
show that the former appears earlier than the latter; and there is no reason 
to suppose that this succession is altered in natural conditions. Hence, if an 
high percentage of the seeds are attacked by /errisia the possibility for Zory- 
mus of finding a sufficient supply of seeds decreases and its number is reduced. 
At the same time the propagation of the normal parasite of Perrisia strobi, 
Platygaster contorticornis, has, presumably on account of-climatic conditions, 
been checked. When therefore, ÅAprostocetus appears, it finds a great number 
of seeds attacked by Perrisia-larvae. which are not parasitized, but only few 
Toryvmaus and, as a consequence, it is forced to attack Perrisza instead of 7o0- 
VNUUS. 
ÅA new method of ascertaining the parasites of the respective 
host-insects. 
The cones being as a rule inhabited by many different insects, it was ob- 
viously impossible to draw any safe conclusions as to the relation of one 
species to another. It became necessary, therefore, to find some method of 
securing data regarding the relation between the noxious forms and the 
parasites. The method employed was to collect daily the insects hatched in 
