230 



W. V. I.AMHKRT AND C. W. KXOX. 



and were tabulated without regard to the week in which they 

 were obtained. On a few of the hens only data from the dead 

 embryos were included, whereas, in others both data from dead 

 embryos and chicks were included. In all cases, however, where 

 the sex ratio of chicks was determined it represented a total hatch 

 of those particular hens for that period. In other words it repre- 

 sents the complete record of the hens for any one interval. These 

 unclassified data appear in Table III. 



TABLE III. 



THE OBSERVED SEX RATIO IN SOME UNCLASSIFIED DATA OBTAINED DURING THE 

 NORMAL HATCHING SEASONS OF 1924 AND 1925. 



The large number of chicks dead between the eighteenth and 

 twenty-first days of incubation appearing in the 1924 data is due 

 to the fact, as mentioned before, that in some cases only the dead 

 embryos of certain hens were sexed. 



In this as in the preceding tables no very definite trend to the 

 sex ratio is apparent. While there seems to be an excess of 

 males in the DS column and an excess of females in the DS column 

 the numbers are probably not large enough to indicate a selective 

 mortality at the different stages of incubation. This would 

 appear especially true since no similar condition appears for the 

 more extensive data shown in Table II. In the chicks hatching 

 listed in Table III. a high sex ratio occurred in 1924 and a low 

 sex ratio in 1925. Hence it would appear that the variations 

 noted in this table are probably due to the small numbers. 



From the methodological standpoint Pearl (1917) has indicated 

 the advantage of basing the sex ratio on families of ten or more 

 individuals. He has compared the mean sex ratio of families 

 of various sizes and finds the following ratios in each : 



