EXPEEIMENTS ON PROTECTIVE COLORATION 463 



Experiment 6. My notes regarding this experiment are not entirely 

 clear in respect to the number of crickets. Either two or four (prob- 

 ably two) were placed on A and C and four on B. The crow twice 

 dropped from its perch to the floor near the backgrounds, and returned 

 to its perch. A third time it flew to the side of the cage nearest A, 

 passed and repassed the three backgrounds and then returned to its 

 perch. A fourth time it left its perch passing to the side of the cage 

 near A, then to a point near C. It then returned to A and took one 

 cricket therefrom. Time of experiment six minutes. 



Experunent 7. Same arrangement of background and prey as in 

 Experiment 6. The crow left its perch almost immediately, passed 

 the backgrounds from side to side twice, and then took one cricket 

 from B. 



Experiment 8. Ten crickets were placed on both B and C. The 

 crow left its perch, passed the backgrounds two or three times, and 

 then took one cricket from B. 



Experiment 9. Same arrangement of background and prey as in 

 Experiment 8. The crow at once left its perch and passed to the side 

 of cage nearest A. It then approached the backgrounds, retreated 

 a step or two, approached again and took one cricket from B. 



Summary 



In this series the arrangement showing the greater contrast 

 was chosen five out of nine times, or 55 per cent. 



SERIES III 



Plate 2, figures 50 and 52 



Experiments with the Purple Martin (Progne subis) and grass- 

 hoppers (Melanoplus^ and Xyphidium fasciatum). 



A pair of purple martins were taken from the nest just be- 

 fore learning to fly. They were kept in a cage 2.4 x 2.4 x 0.8 

 m. in size. At first the birds were very wild, refusing to eat 

 unless forced to do so. They soon became tame enough to eat 

 readily from the hand, however, but it was several days after 

 their capture before they learned to feed themselves. 



^ I have not attempted to differentiate the spp. of Melanoplus used in this 

 and succeeding series of experiments. Both M. atlantis and infantilis are com- 

 mon in the region and these are probably the two mostlj' used. The differences 

 in color and markings between different species are so small as to be negligible 

 for the purpose of these experiments. 



