EXPERIMENTS ON PROTECTIVE COLORATION 491 



by the smallness of the cages in which the birds were confined. 

 Thus instead of being obhged to seek their prey from a con- 

 siderable distance as is the case with many wild birds, notably 

 hawks, it was brought close to the caged birds, thereby reducing 

 very materially the protective coloration effect. 



I have made several attempts to perform similar experiments 

 with wild birds, but in most instances without success. Wild 

 birds ordinarily have an abundance of food available, so that 

 they will pay no attention to food prepared for them, especially 

 if the food be dead, while their natural food is ahve. It is highly 

 probable moreover that movement of their prey plays a large 

 role in enabling raptorial and insectivorous birds to secure it. 

 To this latter point I shall refer later. 



In a few cases, however, I secured results which will be re- 

 corded in the following experiments. 



SERIES X 



Experiment 1. Figures 29 and 50. In this experiment three back- 

 grounds were arranged as shown in the accompanying diagram (fig. G), 

 1 and 3 being composed of blades of grass and 2 of flour. Five Mel- 

 anoplus were placed on 1 and 3 and ten on 2. The latter combination 

 presented the greater contrast. About 45 cm. from 1 and 8 to 10. 

 cm. above the level of the backgrounds was a perch, A. Two minutes 

 after arranging the Ijackgrounds two English sparrows (Passer domesti- 

 cus) lit near them and fed from each leaving three insects on 1, seven 

 on 2 and one on 3. Seven minutes later a kingbird (Tyrannus tyrannus) 

 lit on the perch and then flew over 1 io 2 where it took one insect. 

 It then took three from 1 and two more from 2. When viewed from 

 the perch then at a distance of 46 cm. the color of the insects was pro- 

 tective, but at closer range this effect was lost. 



Experiinent 2. Figures 29 and 32. In this experiment two back- 

 grounds were arranged in the same position as in Experiment 1, 1 

 being composed of a mixture of straw and dry ditch grass (Ruppia 

 maritima) and 2 of sifted flour, on each of which were placed seven 

 Melanoplus. The latter combination presented the greater con- 

 trast. A grackle (Quiscalus quiscula anaeus) soon fed from each back- 

 ground feeding from that one (2) which it first approached, from which 

 it took one, and then two from one of the others. A kingl)ird then 

 flew over both backgrounds and lit on the perch. It then flew over 1 

 and took an insect from 2. 



Experiment 3. The same as Experiment 2 except that five insects 

 were placed on each of the two backgrounds. After a sparrow had 

 alighted and fed on 2, a kingbird flew over both backgrounds to alight 



