138 



Maine; agricuIvTurai, experiment station. 1909. 



same birds' eggs infertile in the second breeding year. We 

 have available complete records for two successive hatching 

 seasons of 52 birds. The correlation tables for per cent, of 

 infertile eggs and per cent, of fertile eggs hatched for these 

 birds are given in Tables XVII and XVIII. 



From these tables the variation constants given in Table XIX 

 have been calculated. 



TABLE XIX. 



Constants for First and Second Years' Hatching Records of 

 the Same Birds. 



Constant. 



Per cent, 

 infertile. 



Per cent, fertile 

 eggs hatched. 



Pullet year mean 



" " standard deviation 

 Second year mean 



" " standard deviation 

 Coefficient of correlation 



16.83±1.28 

 13.66±0.90 

 14. 42 ±2. 00 

 21.37±1.41 

 -0.in±0.092 



49. 81 ±2. 03 

 21.67±1.43 

 51.63+2.29 



24.46±1.62 

 0.331±0.083 



From these tables the following points are to be noted : 



1. There is comparatively little difference in the mean fer- 

 tility or mean hatching quality of the eggs of this group of 

 birds in the two years, so far as the data enable any conclusion 

 to be drawn. This result would seem to indicate that the sup- 

 posed superiority of hens over pullets in breeding performance 

 arises in the main from the fact that the hens kept as breeders 

 the second year are usually selected, consciously or uncon- 

 sciously with regard to their first year breeding records. An 

 average improvement of about two per cent, such as is shown 

 by this group of birds, is certainly not indicative of any marked 

 tendency for a bird to be a better breeder in her second year 

 than in her first. 



2. The variability, both in regard to fertility and hatch- 

 ing quality of eggs is absolutely anc^ relatively greater in 

 the second year than in the first. 



3. Having regard to the magnitude of its probable error the 

 correlation in respect to the fertility of eggs in first and second 

 year is probably to be regarded as not significant. In other 

 words it would appear that, so far as may be judged by the 



