INHERITANCE OF FECUNDITY 257 



will equally well or better account for the facts. If so, by all 

 means let us have it. But in the meantime, it may be fairly be 

 said, that the hypothesis here presented brings together under a 

 few, symbolically simple, general statements a wide range of very 

 diverse and complex facts of inheritance. 



The strongest general evidence that the Mendelian hypothesis 

 here presented is at least a close approximation to the truth in 

 respect to the inheritance of fecundity in the fowl is found in 

 the fact that it accounts equally well for so wide a range of diverse 

 phenomena. In the two 'pure' parent races, one of generally 

 high and the other of generally low fecundity; the two reciprocal 

 •crosses; and the twelve different kinds of matings in F^, we have a 

 series of really independent measures of the validity of the hypothe- 

 sis. It accords with the facts in all but one (the matings of C.I. G. 

 c^ 578 with Barred Fi 9 9 ) of all of the different types of matings 

 tested. The one exception probably has a physiological explana- 

 tion (pp. 246-250). In view of these facts the cumulative proba- 

 bility that the hypothesis applied represents at least a reasonable 

 approximation to the true interpretation of the results becomes 

 very great. 



A possible criticism of the whole method of this investigation 

 might be found in reference to the measure of fecundity which 

 has been used throughout, namely, the winter egg production. 

 Regarding this matter it should be said that the very reason why 

 winter egg production was adopted as the unit of measure in all 

 of the fecundity work of this laboratory was because a thorough 

 biometrical and physiological study of egg production in fowls 

 showed beyond question that winter production was the best 

 practicable index or measure of a fowl's innate or constitutional 

 capacity in respect of fecundity. The reasons for this conclusion 

 have been set forth in this and former papers from the laboratory 

 and need not be repeated in extenso here. The most significant 

 of them is that the differences in observed production between 

 individuals of different innate fecundity capacities are relatively 

 greater in respect of winter productions than of any other time 

 unit that can practically be employed in the measuring of this 

 character. To suppose, however, that the results set forth in this 



