152 THE BEHAVIOR OF UNIT 



in crosses with any other type; it is recessive in 

 the Mendehan sense with reference to all the 

 others. 



Darwin explained the reversion of feral rab- 

 bits to the gray type on two grounds: (1) "a 

 tendency in all crossed animals to revert to their 

 primordial state," and (2) the action of a more 

 " natural " environment when the animals are 

 free than when they are in captivity. In reality 

 neither of these conjectured reasons has anything 

 to do with the case. Some varieties will undei^ no 

 circumstances give reversion, if crossed with each 

 other; but reversion may be obtained as readily 

 in captivity as anywhere. Reversion is due solely 

 to the bringing together of certain unit char- 

 acters, whose joint action is necessary to produce 

 the observed result. 



In producing the gray coat characteristic of 

 wild rabbits at least eight independent unit char- 

 acters are involved. Other color varieties of the 

 rabbit have arisen by regressive variation, i.e. by 

 loss, more or less complete, of one or more of 

 these unit characters. 



To illustrate the matter, let us consider the 

 result of a particular experiment. Black rab- 

 bits were crossed with light yellow (cream) 

 ones, and produced wild-colored gray offspring. 

 These bred inter se produced young of a variety 

 of colors, but among them grays again predom- 

 inated. All the first generation grays seemed 

 to breed ahke, producing young of various colors. 



