Il8 MAINE AGRICULTURAL EXPERIMENT STATION. 1909. 



lation were perfect high fertility would invariably denote high 

 hatching quality and vice versa. The relationship actually 

 exhibited in the 1908 records is about half way between such 

 perfect correlation, and the entire absence of correlation in 

 which event the two phenomena are not in any way related. 

 The 1909 records show a considerably smaller correlation than 

 those for 1908. Here, however, the coefficient for the com- 

 bined data is approximately 2.6 times its probable error and 

 hence would have to be regarded, even when considered by 

 itself as almost certainly significant. When the 1909 pullets 

 and yearling hens are treated sej^arately the coefficients are 

 slightly smaller than when they are combined, and with the 

 reduction in the number of entries in the tables the probable 

 errors are increased in value. In both of these cases, however, 

 the sign of the coefficient remains negative. Neither of these 

 coefficients (1909 pullets only and yearling hens only) could 

 be considered certainly significant in comparison with its prob- 

 able error when taken by itself. Taking both years together 

 there is no doubt as to the conclusion that, so far as the present 

 data indicate, there is a small but still sensible correlation 

 between fertility and hatching quality of eggs. In the long run, 

 or on the average, it is to be expected on the basis of this result 

 that if a hen under a given set of conditions produces eggs high 

 in fertility the fertile eggs zvill also run high in hatching quality, 

 and vice versa. It is to be understood that no wider generality 

 is claimed for this conclusion than arises from the data on 

 which it is based. 



It is not unlikely that the absolute degree of the correlation 

 between fertility and hatching quality of eggs may be different 

 for dififerent breeds. The present data show that this correla- 

 tion is different for different conditions of housing, treatment, 

 etc. The further analysis of the precise effect of these factor - 

 present interesting problems for further work in this connection. 

 It seems unlikely, however, that under any circumstances the 

 correlation would be turned about so that high fertility was 

 associated regularly with low hatching quality and vice versa. 



