114 -MAINE AGRICULTURAL EXPERIMENT STATION. I909. 



TABLE II. 



Shozi'ing the Correlation betzveen Fertility and HateJiing Quality 



of Eggs. Records for Hatching Season of ipo8. 



bation or its general average result has nothing whatever to do, 

 provided that large numbers of eggs are dealt with and all are 

 incubated by the same method. It is with the average differ- 

 ences in the fertility and hatching of the eggs of individual hens 

 all treated in the same way that we have to do. 



From the way in which our records are taken it is easiest to 

 use as a measure of fertility the percentage of infertile eggs 

 in the total number of eggs set and as the measure of hatching 

 quality the percentage of fertile eggs which were hatched. 

 Now in order to determine the correlation between these two 

 variables, per cent, of eggs infertile on the one hand and per 

 cent, of fertile eggs hatched on the other hand, it is necessary 

 to prepare a table which shall show for each bird included in 

 our records the performance in respect to each of these varia- 

 bles. Having prepared such a table it is possible by the appli- 

 cation of proper mathematical methods to deduce an exact 

 numerical measure of the degree of . correlation exhibited.* 



* For references to tlic literature describing these mathematical 

 methods see Me. Agr. Expt. Stat. Bulletin 166, p. 64. 



