Jl6 MAINE AGRICUIvTURAI, EXPERIMENT STATION. I909. 



TABLE IV. 



■Showing the Correlation hetzveen Fertility and Hatching Quality 

 of Eggs. Records for Hatching Season of igog. Pullets 

 only. 



■ 



I 







10 



T 

 















in 



I 



in 







'ee 







T 



in 



CO 



c 



in 

 1 

 5 



ENT. ' 



c in 

 in in 

 1 1 

 10 

 ^ in 



NF 







in 



in 



EKTILE. 



in 10 



T ^ ^ 



in 

 CD CO t~ 







00 



ml 



in 







in 

 00 



in 



I 



05 





 







i 0- 5 



K 5- 10 

 g 10- 15 

 < 15- 20 

 W 20- 25 

 « 25- 30 

 30- 35 

 M 35- 40 

 Z 40- 45 

 g 45- 50 

 P 50- 55 

 ti 55-60 

 fa 60- 65 



65- 70 

 70- 75 



75- 80 

 ^ 80- 85 

 a 85- 90 

 . 90- 95 

 P3 95-100 

 H 100-105 

 Ph 



2 

 1 



2 



1 



5 

 1 

 4 

 4 

 4 

 6 

 6 

 1 

 3 

 4 

 1 



1 



1 

 1 

 1 



1 



1 



2 



1 



1 



1 



1 

 2 



1 

 1 

 1 



1 

 2 



1 

 1 



1 

 1 



2 

 1 



1 



1 

 1 



1 



1 



1 

 1 



1 



1 



1 



- 



- 



_ 



1 

 1 



- 



1 



- 



- 



- 



1 



1 



8 

 3 

 2 

 3 

 2 

 1 

 6 

 4 

 8 

 4 

 9 

 7 

 9 

 3 

 4 

 8 

 4 

 

 

 1 

 1 



Totals 



46 



8 



8 



5 



5 



3 



5 



1 



1 















2 







1 















1 



1 



87 



individuals in those rows fall towards the right hand side of the 

 table. That is, these are birds hatching relatively high per- 

 centages of fertile eggs. But, at the same time, these first 

 three rows denote a relatively low percentage of infertile eggs 

 or, stated the other way, denote that a high percentage of the 

 eggs were fertile. While the existence of this general trend 

 showing a correlation between fertility and hatching quality is 

 evident on inspection, the degree or amount of the trend cannot 

 "be so told. What is necessary is to get some single constant 

 which shall in one figure give a measure of the general trend 

 exhibited by the table. Such a constant is afforded in the 

 so-called coefficient of correlation which can be evaluated from 

 the table by appropriate methods. Calculating this coefficient 

 from each of the two tables and denoting it by the usual symbol 

 r we have : 



