176 BULLETIN 19 7, UNITED STATES NATIONAL MUSEUM 



and while about a foot in the air, tlie slirilie released its hold on the neck of the 

 prey, and, without hesitating or altering its course, caught the sparrow in its feet. 

 The flight was continued for about fifteen yards, and then the shrilie dropped 

 to the ground. It started off at once and the same behavior was repeated ; the 

 prey was picked up by the neck with tlie beak and tliis hold was given up, while 

 flying, for the hawk hold. The substitution is almost instantaneous ; the burden 

 does not drop perceptibly and tlie flight is continuous and steady. 



This observation on behavior, although entirely reliable, must be 

 considered as abnormal, for most shrikes certainly do not handle their 

 prey in this fashion. 



With large-sized prey, shrikes resort to special impaling stations 

 within the territory. The variety of impaling devices is great and 

 indicates a generalization of the inherited reaction, permitting indi- 

 vidual adaptability in the matter of finding and experimenting with 

 the most effective means for the fixation of food. The first attempts 

 of juveniles at impaling have been noted in cage birds. These consist 

 of dragging the food along the perches with a jerking motion, the head 

 held low and the long axis of the bill paralleling the perch. If nails 

 are provided in tlie perches, the food on encountering these obstacles 

 offers resistance to the bird, which continues to tug and may on occasion 

 thus firmly lodge the food on the nails. Subsequently, small bits are 

 picked from the impaled mass, first gently, but later they are torn 

 loose with great vigor. 



Well-known examples of trees and plants used for impaling are 

 orange, yucca, saltbush, and some cactuses. The use of barbed wire 

 is widespread wherever this is available. Ends of broken branches 

 also are used extensively. I have observed a California shrike slip 

 the abdomen of a Stenopelmatus over a broken willow twig one-fourth 

 of an inch in diameter, fitting it on tightly, and then proceed to tear 

 oft' bits from the open anterior end of the abdomen. My cage birds 

 became accustomed to impaling upon the split end of a small broken 

 branch as well as upon nails driven through their perches. 



If shrikes are hungry wdien large prey is impaled, they eat enough 

 to gorge themselves thoroughly, xis much a 7 grams of food may be 

 eaten at a single feeding. If the impaled food amounts to more 

 than this in bulk, the remainder may be left hanging. The shrike 

 returns to the food during succeeding hours and usually eats it to the 

 last morsel. iVs a storing device, impaling is useful for periods of 

 a day or two onlj', unless climatic conditions favor the preservation 

 of the animal food for longer intervals of time. Impaled objects are 

 comparable to the kills of large predatory mammals, and, so long as 

 they are still edible, ih^ shrike returns to them unless more recent kills 

 occupy its attention. Storage in its true form involves gathering to- 

 gether a considerable quantity of food for future use without at the 

 time partaking of it extensively. The object of the American shrikes' 

 habit of impaling, then, is not truly storage. When concentration of 



