NATURAL SELECTION 



selection ) can emerge minus its . . . warning colors, its mimicry and its 

 signal colors, the reaction over the end of the century will have been a 

 distinct advantage." 



Resolution of Difficulties. Meanwhile, data have accumulated which 

 answer satisfactorily the main objections to the theory of adaptive colora- 

 tion. The first objection above was that predators might depend upon 

 utraviolet or infrared vision, and hence that coloration as seen by man is 

 irrelevant. This has been investigated extensively, both by testing the 

 vision of predators in different portions of the spectrum, and by photo- 

 graphing presumed protected animals in ultraviolet and infrared rays, as 

 well as with visible light. The results have shown that any part of the 

 spectrum is used by some predator. Some animals which appear protec- 

 tively colored to the human eye are not protectively colored when photo- 

 graphed by the type of light used by their usual enemies, but many are. 

 And it is entirely possible, indeed probable, that many organisms which 

 we do not suspect of being protectively colored would appear so if photo- 

 graphed at the appropriate wave length. This argument is plainly a 

 two-edged sword. 



The second objection was based upon the lack of proof that presumed 

 distasteful aposematic animals, particularly insects, actually are distaste- 

 ful to predators. In many specific instances, it has been proved that the 

 animals in question actually are distasteful to some, at least, of their 

 potential enemies. For example, when a large variety of freshly killed 

 insects were placed on a feeding tray at the edge of a natural woods, birds 

 were observed to come and feed. They took most of the cryptically colored 

 insects, but very few of the aposematically colored ones. The recent experi- 

 ments of Brower prove that some aposematic butterflies are, in fact, dis- 

 tasteful, for jays which had been conditioned by feeding on models 

 rejected both models and mimics, while unconditioned birds took the 

 mimics readily. Again, the tectibranch mollusc Oscanius membranaceus, 

 which is aposematic, secretes dilute sulfuric acid. It is not taken by most 

 fishes, and otherwise edible materials are refused by most fishes if treated 

 with dilute sulfuric acid. When larvae of the magpie moth. Abraxas gros- 

 sulariata, were offered to lizards and frogs, they were at once seized, then 

 dropped, and refused on all future occasions. The predators then sat with 

 mouths agape, rolling their tongues as though they were trying to get rid 

 of an obnoxious taste. Similar experiments have demonstrated that mimics 

 actually do, in some cases at least, share the protection of their models. 



Finally, there is the important objection, based primarily upon McAtee's 

 study, that protected as well as unprotected organisms are taken by 

 predators in numbers proportional to their populations, and hence that 

 there is no protective value demonstrable. McAtee's study has been seri- 

 ously criticized because of the misleading method of presenting his data. 

 The actual numbers of protected and unprotected insects taken were not 

 recorded. Rather, the numbers of stomachs in which each was found was 

 recorded. Thus, if 1000 specimens of a particular species of insect were 

 found in a total of 80 stomachs, and if 100 specimens of another species 

 were also found in a total of 80 stomachs, the published figures would be 



247 



