impalement even when the bird is not hungry. This food usually is not eaten later. Thus the shrike's "butcher shop" is not primarily a store of food, even though it sometimes serves as such when in times of scarcity remains of old meals are eaten. It is not a gathering of food in time of plenty and saving it for a later use. Rather the placing of many items in one tree is the result of its being a favorable impaling place. And the impaling is behavior developed to overcome the weakness of the claws in a bird whose disposition and strong beak enable it to prey habitually on larger animals which otherwise it could not tear to pieces and eat. BIRD FLAVORS [Illustration] Particularly in the study of insects it has been shown that bright or contrasting and conspicuous colors tend to be associated with ill-flavor in the animals that wear them, while insects with a good flavor tend to be so colored that they are difficult to see. The first is a warning coloration--advertising to a predator that he will not enjoy eating this insect and better leave it alone; the other is concealing color, its function apparently to keep predators from finding their prey. The tasters in the experiments that have been used to work out the above generalizations were usually birds, but, as checks, a variety of other animals were used, and the magpie moth (_Abraxas grossulariata_), for instance, was found to be distasteful to certain spiders, frogs, lizards, various birds, a bat, and finally "the late Dr. Hans Gadow (one of the leading ornithologists of his day), who made a practice of sampling caterpillars, remarked on trying an _A. grossulariata_ that it was quite one of the worst he had ever eaten!" Apparently ideas in taste are similar throughout large sections of the predatory animal world. Reversing the usual role, and using insects (hornets) as tasters of bird flesh, the celebrated British naturalist, Dr. H. B. Cott, has recently studied the question of the palatability in birds and their coloration. Naturally Dr. Cott, with his customary thoroughness, compared hornets as tasters with other animals, including cats and men, and found a surprisingly close agreement in the results. The experimental procedure was to expose the flesh of two different birds (without feathers) at the same time, and see which the wasps ate first. Thus a graded series was built up of the 38 species of birds tested, with a palatability rating of from 1 to 38. The wryneck and the crested lark stood at the top of the list, and the pied kingfisher and the white-rumped black chat, as the least palatable, at the bottom with Numbers 37 and 38. Then, surveying the coloration of the birds, and their habits, Dr. Cott made the important correlation that in general the birds whose flesh was most edible were protectively colored, and those whose flesh was least palatable tended to be conspicuous in color and behavior!