296 DAVENPORT— NEW VIEWS ABOUT REVERSION. [April 23. 



pears in one sex only and not until late in life, we get the relations 

 between the calculated and observed proportions in 86 offspring 

 shown in Table lY. Here again the observed distribution agrees 

 with the hypothesis. 



Table IV. 



Black and Black and Red. Buff. White. 



Expected 65 16 5 



Observed 61 17 8 



We may conclude, then, that reversion is not due to the act of 

 crossing in and of itself. It is rather due to the restoration of the 

 ancestral factors, neither more nor fewer. The ancestral combina- 

 tion occurs in by no means all of the hybrids but only in a pre- 

 dictable proportion. 



In conclusion, a suggestion may be offered as to the frequently 

 observed fact of reversion to the ancestral coloration of domestic 

 fowl which have become feral, as, e. g., in the Hawaiian Islands. 

 Darwin suggests that at least part of the reversion of the domesti- 

 cated animals that have run wild must be due to the new conditions 

 of life. One observation that I have made sheds some light on the 

 question. When domestic races run wild much intermingling of 

 races occurs and the primitive type of coloration, among others, 

 recurs. When, as is usually the case, this is better fitted to survive 

 because of inconspicuousness or other advantage, it will tend to 

 escape the general slaughter. That selective elimination is real in 

 poultry is shown by an experience of mine. In an open field about 

 300 chicks ran at large. About 40 per cent, of them were white, 

 40 per cent, black and 20 per cent, penciled or striped. Twenty- 

 four were killed by crows — all either solid white or solid black 

 except one that was coarsely mottled gray and buff. It is obvious 

 that the self-colored poultry are at a disadvantage because of their 

 greater conspicuousness, and they must be, in time, eliminated, leav- 

 ing only the striped or penciled birds — those of the ancestral type of 

 color. Reversion is here not a germinal but an environmental 

 result, due, however, not to climate but to organic enemies. 



