266 RAYMOND PEARL 



The hatching weight of a bird is not directly comparable with 

 the birth weight of a mammal physiologically. Probably the 

 chief factor in determining the weight of a chick at hatching is 

 the size of the egg from which it hatched. There is a high 

 positive correlation between the initial weight of the whole egg 

 and the body weight of the chick which hatches from it. Fur- 

 thermore, external conditions during incubation, especially with 

 reference to moisture, influence the weight at hatching. Finally 

 the innate constitutional vigor of the embryo, reflected in its 

 ability to grow, is a factor in the determination of the hatching 

 weight. In experimental work which involves hatching weight 

 as a datum, the differential effects of egg size, and conditions of 

 incubation, can be practically eliminated provided the experi- 

 ments are carefully and critically planned and carried out. The 

 innate factor then comes forth in full importance. Such condi- 

 tions are believed to have been reahzed in the present work. 



The frequency distributions for hatching weight are given in 

 table 9, and the variation constants deduced from them in table 

 10. The material in the alcohol series is treated in the two 

 groups "Treated & X Untreated 9 9" and ''Treated d" X 

 Treated 9 9" rather than by individual matings, or by sub- 

 stance. An examination of the hatching weight distributions 

 by substance (ethyl alcohol, methyl alcohol, and ether) shows 

 that there is no essential difference between the different groups. 

 On this account there seems no reason to take the space here 

 which would be required to show all the separate distributions 

 in detail. As a control for these and all other growth data we 

 have the 1913 normals of the same cross, from the same stock. 

 In order that no possible error may creep in from differences in 

 date of hatching, use has been made here, and in all other growth 

 data in this paper, of the normal cross-breds hatching in the 

 month of April, 1913, only. A reference to table 4 will show 

 that the mean dates of hatching for the control chicks used in 

 this study and the chick of the treated series are almost exactly 

 identical. 



