INHERITANCE OF WINTER EGG PRODUCTION 111 



introduce a slight error in the comparisons. The error cannot 

 be estimated, because two lots of birds, hatched a week apart 

 may differ by as much as 30 per cent in average egg production. 

 However the regressions from the mothers' families' average is 

 about the same proportionally as Pearl observed in the case of 

 the reciprocal cross, viz.. Barred Rock males by Cornish females. 



Several cross breds laid more eggs than their own mothers. 

 This happened in two out of the five families, with an average 

 excess of best daughter over mother of 21^ eggs against 15| eggs 

 of mother over best daughter in the other three families. 



If the Cornish are not restricted to the three gametic classes, 

 viz., 5, 6, 9, as is done by Pearl, it becomes possible on Pearl's 

 theory, to secure a set of ratios that fit the observed fairly well, 

 if it be assumed that the zero producers are physiological zeros. 

 This assumption is plausible, because most of the Cornish are 

 late maturing. Without this assumption it is impossible to find 

 theoretical ratios in Pearl's theory that fit the observed ratios. 

 Nevertheless, it will be observed from table 8 that considerable 

 violence is done to the observed ratios in the case of one Cornish 

 female, when an observed ratio of zero high to three mediocre to 

 two zero producers is referred to a theoretical ratio of 2^ high, 

 2^ mediocre, and no zero producers! It is also necessary in 

 other years to make similar changes in fitting observed to 

 theoretical ratios, on the assumption that male no. 9191 belongs 

 to class III. 



There is still another difficulty to be explained on Pearl's 

 theory, and that is the occurrence of a single Cornish female that 

 is unquestionably a high producer. It is difficult to believe 

 that she is a mutation because she is the daughter of a bird that 

 though late hatched and somewhat slow about maturing, laid 27 

 eggs in a single laying period before March 1, i.e., has the 

 characteristics of a late maturing high-producing female. 



On the alternative theory, no difficulties are encountered in 

 fitting observed to theoretical ratios. 



Some of Pearl's observed results are in agreement with the 

 results of our cross, particularly the progeny of Cornish male 

 no. 578 (Pearl, '12, p. 373). Here a number of high-producing 



THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 28, NO. 1 



