Body Covering 39 



kept in black tanks and in white until they had reached, 

 respectively, the maximum shade of darkness or of pallor. 

 Then, into a large, light-colored tank, fifty of each kind of 

 fish were placed at the same time as two of the penguins. 

 The dark fish grew somewhat lighter than they had been, 

 but remained very much darker than the pale ones. To the 

 human eye they were conspicuous against the light back- 

 ground, whereas the pale ones were not. The penguins 

 plunged happily about, scooping up little fish at a great rate. 

 The experimenter stood by, watch in hand, hopeful, but 

 quite unable to see how it was all turning out. At the end of 

 two minutes the penguins were removed and the fish counted. 

 Fourteen of the dark fish had been eaten, but only six of the 

 pale ones. A four-hour test was next made, and this time 

 twenty-nine dark fish were eaten to only fourteen pale ones. 

 In a light-colored tank, twice as many dark fish had been 

 eaten as light-colored fish ! Protective coloration had actually 

 protected. Four more tests gave the same answer, although 

 not quite so strikingly. And when the reverse of this experi- 

 ment was tried, putting both kinds of fish into a black tank, 

 the results, perhaps because the dark residents were here even 

 less visible than the pale ones had been in the light tank, 

 were still more convincing: 217 pale victims in all were 

 eaten, against only 78 dark ones — 74 per cent pale to 26 per 

 cent dark, almost three to one. 



Not content with this, the experimenter went on to try 

 his system on other predators. He used a heron, which sees 

 its prey only from above, and got approximately the same 

 results. And then, inasmuch as the majority of predators of 

 fish are other fish, he used a species of sunfish which hap- 

 pened to be available, and got even more definite results. 

 After years of futile theoretical argument, one man had 

 solved the old problem by the simplest and most straight- 

 forward of methods. To the satisfaction of most people, he 



