INHERITANCE OF WINTER EGG PRODUCTION 89 



will make them available to all and at that time to consider the 

 subject from several other angles, among them that of selection. 

 We may anticipate the proposed work sufficiently to note that 

 egg production yields readily to selection on a suitable basis, as 

 Pearl ('15 a and b) and Dry den ('16) have already found. 



The position of the numerical division point between high and 



mediocre producers 



The determination of a suitable numerical division point has 

 been a matter of concern in our work, for it represents the line 

 of separation between two genotypes, which, however, may 

 overlap. The obvious method of finding this point is to plot a 

 distribution curve of the winter egg production, and determine 

 the point at which the antimode occurs. This method indicates 

 no other division point for our Rhode Island Reds except the 

 one at or near 30 eggs. At the same time the evidence is not of 

 a very striking character. Further, I have shown that the time 

 of hatching has a marked influence on the shape of these curves 

 (Goodale, '18). We have, therefore, assumed other division 

 points and worked out a comparison between the observed and 

 theoretical ratios which is given in table 5. 



A division point at 40 eggs has some evidence in its favor. 

 An inspection of our records indicates that February 15 corre- 

 sponds more closely to the end of the winter period than March 1, 

 because during the last half of February there is a resumption 

 of production on the part of birds that have undergone a winter 

 pause. Moreover, the curve of daily egg production shows a 

 sharp upward turn at this time. The rate of production of birds 

 at this season is sufficiently high to make the last fourteen days 

 of February yield 9 or 10 eggs, so that an egg production of 40 

 eggs before March 1 is the equivalent of 30 eggs laid before 

 February 15. Because of the form in which the records are 

 transcribed, it is much simpler to handle the data with a division 

 point at 40 eggs laid before March 1 than to use a division 

 point at 30 eggs laid before February 15. There is no objection 

 to this method of treating the data, for the very few bu'ds that 



