Summary..(1) To test the hostility reactions of Black‐headed Gulls(Larus r. ridibundus)the following were employed:–Stuffed mounts, one in summer and one in winter plumage; a dead bird in summer plumage, propped up; the same bird decapitated; and its isolated head.(2) Hostility reactions were immediately elicited by both the mounts, provided they were within a certain distance from a nest. This territorial radius averaged about 3‘ 6“, varied with the temperament of individual birds, and with the temporary state of one and the same bird. No hostility was shown by occupants of more distant nests, nor by any birds when the mount was set up on neutral ground away from the nesting area.(3) Hostility was also evinced towards the dead bird, but, presumably owing to its less natural pose, the reaction was less extreme.(4) In all these cases hostility was chiefly shown by pecking the back of the head. The feet were also used while flying over to attack the back, and occasionally other parts than the head were pecked. It is uncertain whether the head was less singled out for attack in the winter than in the summer plumage mount.(5) No attacks were made upon the decapitated dead bird, in spite of its causing obvious anxiety and preventing return of birds to the nest near which it was placed.(6) The isolated head, when set up on a skewer, was attacked, in this case being pecked on the beak as well as from behind. On one occasion when it fell into a nest with eggs, it was pecked lightly and then treated like an egg and brooded.(7) It is concluded that two factors compete in eliciting hostility:–(a)Resemblance in form and pose to a normal individual; (6) the dark‐brown head of the breeding bird. Either alone is operative, but with both together (summer plumage mount) the result is possibly more intensive, and probably more specific, the dark head acting as a directive for the attack.(8) Suggestions are made for further experiments to test and amplify these