POULTRY WORK. 213 



3 years, it may be said that it clearly appears that these factors 

 of flock size and floor space have a definite and measurable effect 

 on the average annual production. This effect is quite consid- 

 erable in amount. The bearing of the results of this floor space 

 experiment on the breeding experiment with which we are 

 chiefly concerned here lies in the fact that the figures which will 

 be given in detail in the complete paper show beyond any doubt 

 that these environmental factors can, even after the close selec- 

 tion for more than 5 years, still cause very marked changes in 

 the character (egg production) which it was hoped to fix in the 

 strain by breeding. There is a considerable amount of detailed 

 evidence which is presented in the complete paper all of which 

 tends to show that the quality of high productiveness cannot 

 be regarded as any more a fixed characteristic of the Station's 

 strain of Barred Plymouth Rocks now than it was at the begin- 

 ning of the experiment. 



The general conclusion regarding the results of the breeding 

 experiment may be quoted from the complete paper. "The 

 practical conclusion to be drawn from the results of this breed- 

 ing experiment seems to the authors to be clear. It is that the 

 improvement of a strain of hens in egg producing ability by 

 selective breeding is not so simple a matter as it has been sup- 

 posed to be. Nothing could be simpler than breeding from high 

 producers to get high producers. But if this method of breed- 

 ing totally fails to get high producers — in other words, if the 

 daughters prove not to be like the mothers in egg production — 

 it cannot fail to excite wonder as to whether the simplicity of 

 the method is not its chief (possibly its only) recommenda- 

 tion. Anyone who makes a thorough, first-hand study of an 

 extensive selection experiment carried out, as was this one, by 

 the so-called German method without testing of the centgener 

 power of the individual organisms, cannot fail to be impressed, 

 we believe, with the fact that the improvement of a race by 

 selective breeding is a vastly more complicated matter than it is 

 assumed to be by those who maintain that one need only to 

 breed from the best to insure improvement. The supposed 

 "facts" of heredity on which the practical stock breeder (work- 

 ing for utility points) operates are in very large part inferences 

 rather than facts. What is needed more than anything else for 

 the advancement of the stock breeding industry in all its phases 



