BROODINESS AND FECUNDirV IX VOWl. 79 



decline is observed until the group with six broody periods is reached, after 

 which the mean egg yield falls significantly. The probable error increases so 

 rapidly due to small numliers when the class with eight broody periods is 

 reached tliat very little significance can be attached to the mean in this and 

 later classes. On the wliole, this table suggests in a general way that increased 

 liroodiness does lower the annual record. 



SO. Correlation Befzceen Times Broodif and Annual Egij Yield for Broodi/ 

 Birds Alone — Pnllef Year. 



A true measure of the correlation lietween times broody and annual pro- 

 duction can only be found witliin the broody po]iulation as previously stated. 

 In the eight-year period 1122 broody birds are concerned. This group lias 

 been tabulated and constants calculated. 



Number of birds .......] 122 



Mean times broody ...... 2 89 



Times broody standard deviation .... ±1.91 



Mean annual production ..... 164.89 



Annual production standard deviation . . . ±45.03 



Coefficient of correlation —.1791 ± .0195 



Regression broodiness on production . . . — .007() 



Regression production on broodiness . . . — 4-.2167 



A negative coefficient of correlation of .1791 ± .0195 indicates that the 

 correlation between times liroody and annual jiroduction may not be consid- 

 ered intimate. Sucli a constant leads to the assumption that broodiness as 

 measured by periods iias played some part in limiting annual production for 

 the eight-year period studied. 



■U. Correlation Beticeen Total ]>ai/s Broodi/ and Annual Production — Pullet 

 Year. 



Broodiness may next lie measured in total days for the year. Tliis method 

 of measuring has been made use of between total days broody, and annual 

 egg jiroduction. The birds used in this tabulation are 2245 in mmiber, both 

 liroody and non-broody. Constants calculated follow: — 



Ninnlier of birds ....... 2245 



Mean total days broody 23.68 



Total days broody standard deviation . . . ±26.98 



Mean annual production ..... 173.06 



Annual production standard de\ iation . . ±46.40 



Coefficient of correlation —.2200 ± .0135 



An interesting and important fact is brought out by the above constants 

 in that the correlation between total days broody and annual egg record is 

 negative and of almost the same magnitude as that obtained between times 

 liroody and annual egg record (section 29). The deduction that broodiness 

 lowers annual record again seems warranted, but the coefficient o1)tained in 

 this talile is false because of tiie presence of the genetically different broodies 

 and non-liroodies. 



