( fix ) 



an Ichneumon, but the hovering and darting flight suggested 

 rather a Syrphid. It was so extraordinarily active that I 

 failed to catch it. At the time I was puzzled by this, for I 

 couldn't see the object of laying an egg in an insect which was 

 destined to be cut up into little bits at once ! I was therefore 

 extraordinarily pleased, when recently re-reading Bates on 

 the Amazons, to note that he describes a precisely analogous 

 thing in the case of a fly of genus Stylogaster (Conopidae) and 

 the foraging ants Eciton. He says ' the armies of all Ecitons 

 are accompanied by small swarms of a kind of two-winged 

 fly . . . these swarms hover with rapidly vibrating wings 

 at a height of a foot or less from the soil over which the Ecitons 

 are moving, and occasionally one of the flies darts with 

 gi'eat quickness towards the ground. ] found they were not 

 occupied in transfixing ants, although they have a long 

 needle-shaped proboscis, which suggests that conclusion, 

 but most probably in depositing their eggs in the soft bodies 

 of insects which the ants were driving away from their hiding- 

 places. These eggs would hatch after the ants had placed 

 their booty in their hive as food for their young.' 



" Isn't it extraordinarily interesting that two such different 

 .species of ants, but of precisely similar habits, should be 

 attacked by parasites in the same way in South America and 

 Uganda ? If one thinks of it there is very little chance for 

 an enemy to attack these ants, which are so active and ferocious 

 and of wandering habits. So either this method, or the 

 method of stealing the pupae which Lamborn described 

 recently, had been evolved as a means of checking such a 

 formidable species. But what extraordinarily fine adjust- 

 ment to the habits of the ant ! The method of gaining an 

 entrance into the inaccessible nest reminds one rather of old 

 stories such as the wooden horse of Troy, etc. I 



■' I once saw a Hemipterous insect escape being eaten by 

 Dorylus. It was one of the flat, triangular, vegetable-feeding 

 type. The ants were all over the bush and frequently seized 

 an antenna or a leg of the bug, but always let go again. This 

 is interesting, because they will eat such distasteful things a« 

 Acraeine larvae and pupae. 



■■ In almost the first column of Dorylus which I saw on the. 



