82 MASS. EXPERIMENT STATION BULLETIN 264 



probable error. The relation bet-ween the two characteristics being studied 

 was not linear, and therefore the correlation ratio expresses the correlation 

 desired. The X^ test for linearity of regression showed a probability of .0501 

 of linearity and confirmed the Blakeman test. 



10. First-Year Effg Production versus Length of Molt Period. 



First-j^ear egg production records cover a 365-day period beginning with 

 the first pulkt egg. These records were arbitrarily terminated on the 365th 

 day regardless of whether or not the bird was layinig at the time. Such egg 

 records were significantly lower than previous^production records in highly 

 persistent individuals. In general, first-year records are comparable with 

 previous-production records in studying the effects of intense physiological 

 activity of the reproductive system upon duration of molt period. 



The same population of 929 birds considered in Section 9 was used to study 

 the correlation between 365-day production and duration of molt period. 

 Class intervals of 10 were used for both characteristics. The following cons- 

 tants appear in the correlation table: 



Number of birds - - 929 



Mean length of molt period — days. — - 75.60 



Molt period standard deviation — days ±30.38 



Mean first-year production — eggs - 223.99 



Production standard deviation — eggs ±35.1+ 



Coefficient of correlation - —.3010 ± .0201 



Correlation ratio _ .3592 



The mean 366-day production was 224 eggs compared with a mean of 

 231.19 eggs produced previous to molt in the same population. The standard 

 deviation in production was smaller on the 365-day records than on the pre- 

 viousrproduction records, as might be anticipated. 



The correlation between 365-day production and length of molt period was 

 tested for linearity, and the regression of duration of molt on egg production 

 gave a probability of .3031 of linearity by the X^ method. 



The correlation ratio of duration of molt to 365-day production was .3592. 

 By Blakeman's test this constant showed a difference from the correlation 

 coefficient 4.67 times as great as the error of difference. It is probable that 

 the regression was rot linear and that the correlation ratio expresses the 

 true correlation. The correlation between 365-day production and duration 

 of molt was —.3592. It is a significant constant and differs but slightly 

 from the similar constant obtained in Section 9. 



Duration of Molt as a Measure of Vigor and Second- Year Production 



As a measure of vigor Hays and Sanborn (1928) have made use of mortal- 

 ity rates for pullets during the first year of laying. Such mortality rates 

 were calculated for each family and the families may be directly compared 

 in this respect. The following studies were made to discover whether degree 

 of vigor was associated with rapidity of molt. 



Second-year egg production is of considerable economic importance. Birds 

 that fail to lay heavily the second year, even though their pullet record was 



