I>K()()ni\i:SS AMI FF.C rNDll'i IN I'OWl. 71 



XuiiilnM- (if birds ....... 2"J'_M 



.Moan total day.s l)roody ..... 2'.i.'Hf 



'!\)tal days broody standard deviation . ±27.01 



Mean winter rate 6(>.|..5 



Winter rate standard deviation .... ±9.37 



Coeflieient of eorrelation +.0178 ± .0142 



Tlie coeflieient of eorrelation is praetieally the same fi}>iire as was ol)iained 

 between times broody and winter rate. This is also a false eorrelation be- 

 eau.se l>roodies and non-broodies each represent a genetic type. Since winter 

 production and annual production are so intimately correlated (Hervey 1923; 

 Hays, Sanborn and James 1924), high winter record is of very great import- 

 ance in lireedlng for fecundity. 



Blakeman's test for linearity of regression has lieeh api)lied in this studv 

 with the following results: 



Correlation ratio for days broody . . . -)-.1134 



Correlation ratio for winter rate . . . -(-.1695 



(Cor. Ratio)2— (Cor. Coeff.)2=i . . . .007.5 ± .0024 



(Cor. Ratio)2— (Cor. Coeff.)2 = . . .0233 ± .00^3 



The difference between the correlation ratio for winter rate squared and 

 the correlation coefficient squared is .0233 ± .0043, a difference more than five 

 times as great as its probable error. This fact indicates that the coefficient 

 of correlation is false, as might be anticipated from the fact that two genetic 

 races are concerned. 



10. Correlation Between To^nl Dai/s llroo'l;/ and Winter Rale for Broo'l;/ 

 Birrh Alone — Ptillet Year. 



Winter rate records are available on 1098 birds tiiat were broody the first 

 year. In this study days broody is tabulated against winter rate to further 

 disco\er the correlation between degree of broodiness and winter rate. Con- 

 stants are as follows: — 



Xuniber of liirds 1098 



Mean total days broody 42.8.5 



Total days broody standard deviation . . . ±27.14 



Mean winter rate ....... 67.57 



Winter rate standard deviation .... ±9.63 



Coefficient of correlation —.0241 ± .0203 



Regression broodiness on rate .... ^.0679 



Regression rate on broodiness .... — .008.5 



The coefficient of correlation as siiown above signifies independence between 

 degree of broodiness and winter rate. The intensity of tlie broody trait is 

 therefore of no concern in affecting winter intensity. 



20. Correlation Beticeen Total ]>ai/s Hroodii and Annual Rate — Pullet Year. 



The total days broody for each bird are tal)ulated against iier aiuiual rate. 

 The lowest rate class is 16-20; the higiiest rate cla.ss is 86-90 Tlie lowest 

 broody class is 0-9; tlie highest broody class is 150-159 days. This study on 

 the 2245 birds used in section 11 will show if broodiness is an advantage or 

 disadvantage from the standpoint of annual rate. Are broodiness and high 

 intensity linked toeether? Constants calculated are:— 



