476 R. T. YOUNG 



they were so slow to feed is a mystery. It cannot be explained as 

 due to unfamiliarity with the rats as food, for they were accustomed 

 to feed upon them. * I shall refer to this question later (p. 493). 



Experiment 18. This was a repetition of the preceding experiment. 

 For an hour and a half the birds were not observed. At the end of this 

 time I returned to the cage and found a hawk eating the rat on the 

 gypsum. 



Experiment 19. Figures 19 and 21. Two Epimys were placed, 

 venter uppermost, one on ashes mixed with gypsum, earth and clay, 

 and one on moist earth. Both were partly imbedded in the back- 

 grounds so as to bring their bellies flush with the surface of the latter. 

 The rat-earth combination showed the greater contrast of the two. 

 After an hour and twenty minutes returning to the cage I found two 

 hawks fighting over the rat on the earth. 



Summary 



In this series of nineteen experiments, the combination of 

 greater contrast was chosen in sixteen or 84 per cent of the cases 

 and that of less contrast in three or 16 per cent. In one of the 

 latter (Experiment 4) the light was such at the time of experi- 

 ment that objects on a light surface (ashes) were in general 

 much more visible than those on a dark surface (moist earth). 

 In another of these (Experiment 6) it is probable that the hawk 

 did not see the mice until it alighted close to them, at which 

 distance the resemblance between the mouse and the clay could 

 not have had much effect. I shall consider this point further 

 in my discussion of results. In Experiments 1, 2, 3, and 14, 

 also, the weak light enhanced the effect of the white background. 

 If these six experiments (1, 2, 3, 4, 6, and 14), regarding which 

 some uncertainty was introduced by the conditions just men- 

 tioned, be removed from the count, the results stand: Combina- 

 tion showing greater contrast chosen in twelve out of thirteen 

 experiments (92 per cent), that of less contrast in one (8 per 

 cent) . 



SERIES VI 



In this series a long-eared owd (Asio wilsonianus) and house 

 mice (Mus musculus) were employed. The former was taken 

 from the nest in June just before learning to fly and kept in a 

 small cage until October, when the experiments were performed. 



