BIRDS AND INSECTS 173 



pillars by spraying and tar-banding the trees and 

 erecting nesting-boxes for the birds." 



We have then at least independent evidence that 

 the only landowners of the district that adopted re- 

 medial measures did benefit to the extent of at least 

 two years' growth of timber as compared to those who 

 did not. 



Our investigations on the activities of other agencies 

 at work in the control of the larch saw-fly led to some 

 interesting and important discoveries. In addition to 

 the small birds that feed upon the saw-fly and its 

 caterpillars in the summer, we found that the starlings, 

 pheasants, and probably the rooks as well as the 

 field-mouse destroy large numbers of the pupas as 

 they lie on the ground in the winter, but probably 

 the most destructive agent is the ichneumon parasite 

 Mesoleius tenthredinis . In the winter months we 

 collected a large number of the pupae and kept them 

 in cages until the insects emerged, and we found that 

 a certain number were always parasitized, the ichneu- 

 mon emerging from the pupa instead of the saw-fly. 



Thus in 1908 we found that 5.8 per cent of the pupas 

 that hatched were parasitized, 10.9 per cent in 1909, 

 and 62 per cent in 1910. In 1911, however, the per- 

 centage fell to 18, in 1912 to 8, and was even less 

 than that in 1913. But with the diminution in the 

 percentage of pupas parasitized by the Mesoleius, we 

 found an increasing number were parasitized by two 

 other insects, the Ichneumon Hypamblis albopictus 

 and the Tachinid fly Zenillia pexops, and thus the 

 war of insect upon insect never ceased to rage. The 

 important part that is played by these parasites in 

 keeping this pest in check cannot be overestimated. 



It is probable that under normal conditions there 

 are always some larch saw-flies in the larch plantations, 

 but the number of those that survive the attacks of 



