258 RAYMOND PEARL 



paper depend for their existence upon the use of this particular 

 time period of production as a measure of fecundity has no warrant 

 in fact. Precisely the same results (in principle) would be ob- 

 tained if yearly production records were used in the analysis. 

 During the whole of this work complete yearly records have been 

 kept and have been studied. They show in every essential par- 

 ticular the same kind of results as those of this paper. There are 

 objections to the use of the year as a unit of measure, however, 

 which may not be obvious to one inexperienced in these matters. 

 In the first place, it is very much more difficult to keep large flocks 

 of hens in normal, and healthy physiological condition over a 

 whole year period than over a shorter period. Again the risk of* 

 an accident (say the use of bad feed or something of the kind) 

 occurring and upsetting the birds physiologically, and coincidently 

 rendering their fecundity records abnormal and in greater or less 

 degree useless, is increased just in proportion as the time unit is 

 increased. Further the year period includes as a too dominant 

 feature, the spring egg production. The production during the 

 months of March, April and May is practically worthless (and 

 has long been so recognized by experienced poultrymen) as an 

 index or measure of the true, innate or constitutional fecundity 

 capacity of the individual. During these months (in northern 

 latitudes) all hens which are not diseased, malformed, infantile 

 or senile, lay anywhere from 'well' to 'very well.' There is rela- 

 tively little difference between the most and the least fecund at 

 this season. . This period is therefore worthless as a measure of 

 fecundity, and its inclusion in any longer period makes that by 

 so much the less valuable as a measure. 



In view of all these considerations it seems certain that the 

 results obtained are not open to criticism on the ground of the time 

 unit used as a measure of fecundity. 



Another matter which needs careful consideration is as to the 

 possibility of unconscious bias having influenced the results them- 

 selves. In other words, to what extent does the personal equa- 

 tion factor enter into this fecundity work? It can be fairly said, 

 I think, that there is less opportunity for unconscious bias to 

 affect the results here than in genetic work on most other charac- 



