SPECIES. 101 



Uphora vomltoria), and in a single instance, a mourning-cloak butter- 

 fly, it ate at a single gulp, but very large insects, such as tumblebugs 

 {Oopris Carolina), it always impaled. It refused larvse of the mourn- 

 ing-cloak butterfly, the forest tent caterpillar, the fall webworm, and 

 the tussock moth. It would not eat a skin-beetle (Trox), but took 

 with relish May-beetles {Lachnosternd), flower-beetles (Trichius piger)^ 

 and long-horned beetles {Monohammus). Insects provided with espe- 

 cial protective devices were used in some of the experiments. Drone 

 and worker honey bees were eaten, but with no apparent relish. The 

 highly flavored cabbage bug {Murgantia histrionica) was rejected, but 

 its near relative {Eiischistus), a stink bug, was greedily devoured. The 

 investigation of insect food was concluded with tests by means of cer- 

 tain beetles possessing ill-flavored, highly irritating secretions. A 

 burying-beetle {Silpha imequalis), a 12-spotted cucumber-beetle {Dia- 

 hrotica 1^ -punctata), and a blister-beetle {Epicanita vittata), were 

 refused, but an oil-beetle {Meloe wnguaticollis), provided with very 

 powerful irritating secretions, was eaten with relish. The moth of 

 the salt-marsh caterpillar, an insect with a vile odor, was also very 

 palatable. The green ground-beetle ( Calosoma scrutator), yfh\ch throws 

 out a nauseous volatile, acid fluid, highly irritating, was tried. When 

 this big insect was placed in the cage, the shrike seized it by the pro- 

 thorax, bit it vigorously and knocked off its head, and then impaled 

 the body. The beetle's pungent odor filled the room, but the shrike 

 removed the elytra with its bill, and after swallowing the thorax 

 ' bolted the abdomen at a gulp. In the next experiment a Calosoma 

 was so presented to the shrike that the latter had to seize it by the 

 tip of the abdomen. The beetle ejected its irritating ill-smelling secre- 

 tion full in the bird's face. The shrike staggered an instant, then flew 

 to the farther side of the cage, apparently in distress. It was several 

 days before it dared to accept another Calosoma. 



A series of experiments with mice, birds, and other vertebrates was 

 also made. When a live mouse was placed in the cage the shrike gave 

 chase, half running, half flying. It soon caught the animal by the 

 loose skin of 'the back, but quickly let go because the little rodent 

 turned on it savagelj'. In the next attack it seized the mouse by the 

 back of the neck and bit through the skull into the base of the brain, 

 causing instant death. (A broad- winged hawk experimented with at 

 the same time alwaj^s killed its victims with its talons, never touching 

 them with its beak until they, were dead.) A honey -locust perch, set 

 with sharp thorns 2 inches long, had been put into the shrike's cage, 

 and on this it fixed the mouse, a thorn entering below the shoulder 

 blade and passing out through the breast. Then (10 a. m.) it ate the 

 brains. At 10.30 it picked from twentj^ to thirty mouthfuls of hair 

 from the hind quarters, made incisions and removed the skin, and then 

 ate the large muscles. By 11.30 it had devoured the whole bodj% 



