250 THE VARIATION OF ANIMALS IN NATURE 



but McAtee assumes that the number of individuals will be 

 roughly proportional to the number of described species in 

 the group (at least as far as families are concerned), and the 

 number of records from birds' stomachs is compared with the 

 numbers of species recorded in each family for the U.S.A. 



On the whole the correspondence between these numbers 

 is fairly close, though, as might be expected, there are also 

 a good number of discrepancies. We doubt, however, whether 

 McAtee is justified in drawing from his figures the conclusion 

 that all animals are preyed on in proportion to their abundance, 

 and that therefore conspicuously coloured and presumably 

 protected species actually gain no advantage. To substantiate 

 any such far-reaching contention the correspondence would 

 have to be very much more accurate and the results would have 

 to be given in very much more detail. If protective or warn- 

 ing colours have evolved under the guidance of selective 

 predation, we can affirm that the following state of affairs 

 must have existed in the past (and may still exist) : 



That the group in question was attacked by predators. 

 That certain variants were somewhat less attacked than 

 others. 



Obviously such conditions might be fulfilled in a group 

 which, even after a long evolutionary progress, was still very 

 heavily attacked, and McAtee's data throw little light on 

 the problem. 



If the colours are of a mimetic type, then all a selectionist 

 need affirm is that at some time in the past (and possibly also 

 at the present day) more predatory attacks were avoided than 

 encountered by each step in the direction of the model. This 

 again is consistent with a relatively high rate of predatory 

 attacks at the end of the process. In fact, the kind of evidence 

 required to prove or disprove the theory that animal coloration 

 has evolved under the influence of selection is exceedingly 

 difficult to obtain. Although this difficulty may reduce the 

 value of adverse criticism, it is also a distinct drawback to the 

 theory as a whole. 



On the other hand, we believe that McAtee has made a 

 very important contribution, for several reasons. There can 

 be no doubt that the examination of the actual food of pre- 

 dators in nature is the only way of discovering what they feed 



