22 



MASS. EXPERIMENT STATION BULLETIN 438 



At the Massachusetts Station annual persistency has been measured by the 

 number of days from first pullet egg to the cessation of laying within 365 days. 

 Biological persistency has been measured by the number of days from first egg 

 to the cessation of laying associated with annual molt. Of the five types of 

 partial trapnesting, August records should have value for predicting both types 

 of persistency. 



Annual Persistency 



In a population of 1375, the simple correlation between August production 

 and annual persistency was .2907 ± .0167, but regression was non-linear. The 

 correlation ratio was .3473. These values are of considerable significance, and 

 the regression coefiicient has been used for estimating persistency in Table 19. The 

 estimated values agree very well with actual values, indicating that August 

 trapnest records should give valuable information on annual persistency. 



Table 19 — Limited Tr.apnesting as a Mean of Estimating Persistency. 



1 Regression linear. 



2 Regression non-linear. 



Length of Biological Year 



In a population of 515, the correlation between August production and length 

 of the biological year was .2270+ .0282, with linear regression. The magnitude 

 of this constant would indicate an important association. Regression of length 

 of biological year on August production was calculated, and estimated and 

 actual persistency are set down in Table 19. The estimates agree very well with 

 the actual length of the biological year. These data, together with those in the 

 previous section, indicate that August egg records furnish valuable information 

 on persistency. 



