Il8 MAINE AGRICULTURAL EXPERIMENT STATION. I9II. 



matter of size. A trap nest * record was kept of the exact 

 individual egg production of each of these birds. 



RESULTS OE i;aRLIE;R WORK. 



The essential results of the two lines of investigation de- 

 scribed above may be very briefly set forth here. Greater details 

 are given in papers (1-8) cited above. These results are: 



1. That mass selection for high egg production on the basis 

 of the trap nest record of the individual alone did not, as a 

 matter of fact, result in a steady, continuous imprj;)vement in 

 average flock production, though it was continued for a period 

 of ten years. 



2. That, as a matter of fact, the daughters of "200-egg" 

 hens with from 6 to 9 years of mass-selected ancestry (on the 

 basis of trajD nest records) behind them were no better layers 

 on the average than birds bred from the general flock. 



Now whatever opinion anyone may hold as to the biological 

 interpretation of these results he must not after all forget that 

 they are facts. While it has been argued that 10 years is far 

 too short a time to learn anything about the effect of selection 

 it should be remembered that he who makes this argument is 

 really discussing a very complex theoretical matter. An un- 

 biassed examination of the literature on the subject indicates 

 that the length of time which is considered necessary to prove 

 experimentally the effectiveness or non-effectiveness of mass- 

 selection depends almost entirely upon which way the results 

 are coming. If after following a plan of mass selection for 

 even 3 or 4 years one finds that concurrently there has been an 

 improvement in the character selected for, he almost invariably 

 and quite humanly concludes that the selection is the cause of 

 the improvement. Just why, however, 'post Jioc should be con- 

 sidered to be propter hoc when it happens to be "your" hoc but 

 not at all so when it is "my" hoc that is concerned has never 

 been clear to the writer. It certainly seems fair to suppose 

 that it requires just exactly as many years critically to prove by 

 experiment that mass selection in a particular case is effective 



* For a description of the trap nest used in the breeding work of the 

 Station see "Apphances and Methods for Pedigree Poultry Breeding" 

 by R. Pearl and F. M. Surface. Me. Agric. Exp. Station. Bulletin 

 I59> pages 239-274. 1908. 



