EXPERIMENTS ON .PROTECTIVE COLORATION 469 



the notice of the bh-ds more to one background than to another. 

 While the prey in most cases contrasted much more strongly 

 with one background than with the other, it was nevertheless 

 distinct (to my eye) on both at the distance from which the hawks 

 viewed it, the color contrast and resemblance not being very 

 great in any case. In spite of this fact, in a large percentage of 

 the experiments, the prey was taken from the background of 

 greater contrast. In all except Experiments 16 and 17, the direc- 

 tion of the light was either directly or obhquely in the face of the 

 birds. It is noteworthy that in these two experiments, the prey 

 was taken from the background showing less, while in the others 

 it was taken chiefly from the one showing greater contrast. 

 Whether there was any relation between the direction of light 

 and the results of the experiments, or whether the results in 

 Experiments 16 and 17 were merel}^ a coincidence, I cannot say. 



Experiment 1. Figure 2. One Blarina was placed on earth and 

 one on lime. One hawk left its perch and after flying about the cage 

 two or three times dropped to the ground and espying the shrew on 

 the lime, ran directly to it and stood watching it for several minutes, 

 but did not seize it. It apparently did not see the other shrew. In 

 this and the following experiments (up to No. 9 inclusive, and 14, 16, 

 and 17) the greater contrast was presented by the prey on the lime 

 background. 



Experiment 2. Figure 2. Four alternating backgrounds, two each 

 of lime and earth were prepared in this experiment, and one Microtus 

 placed on each. Three hawks on perch. One immediately left the 

 perch, flew to the ground and took one mouse from the lime. 



Experiment 3. Figure 2. Two backgrounds, one of lime, and one 

 of earth, were employed in this experiment, and one Microtus placed 

 on each. Three hawks on perch. For about three-quarters of an 

 hour the mice remained unnoticed by the hawks, although one of 

 the latter flew across the cage once during this interval. I then drove 

 a second hawk from the box, w4iich flew across the cage, di'opped to 

 the ground and ran directly to the mice, but apparently did not see 

 them. Then the first hawk flew to the mice and took one from the 

 earth. ^'^ 



1" It is very possible that this mouse had been disturbed by the second hawk 

 which was standing close to it when it was seized b}^ the first, the attention of the 

 latter being thereby attracted to it. I have often noticed that a hawk will be 

 attracted to its prey, which it has previous!}' apparently not seen, by another 

 hawk seizing it. 



