HKOODIXKSS AM) KKC rxniT^ 1\ |-( )\\ 1 . (53 



Tlu" very liiryv standard doviatioii in tinies ln-oody .su{,''<>:csts a most pro- 

 nouiu'cd varial)ility in mimluM- of l)roody periods. The actual range is from 

 1 to 1'2. Since tlie modal class is broody l)ut once, there can he l)ut little 

 furtlier progress in reducing tiic mean numl)cr of liroody jicriods witlijn tlie 

 broody race. 



'Hie mean December rate i.s slightly higher than that for both broodies and 

 non-broodies combined, in section 1. The standard deviation in rate is of the 

 same magnitude as that in section 1. 



The coefficient of correlation between degree of broodiness and December 

 rate is actually less than its probable error, and since it is of verv small 

 magnitude, the interpretation seems justified that December rate is indepen- 

 dent of degree of broodiness, and that the correlation in section 1 is false. 



• '. Correlation bet'ui'een the Presence of Broodiness and December Utile above 

 the Mean of Broodies and Xon-broodies Combined — Pidlet Year. 



The actual correlation between the presence of broodiness and higii rate 

 is of much importance to the breeder. Such a constant was calculated for 

 tlie 191.5 broody and non-broody birds by the method of Yule (loc. cit.). 



Coefficient of correlation -I-.1466 rt .0150 



Although the degree of correlation between the presence of broodiness and 

 high December rate is not large, there can be no justification for any other 

 deduction than that the presence of broodiness is i>artially linked with high 

 December intensity. The elimination of the broody trait .should result in 

 something of a reduction in December rate for the flock as a whole. 



A furtlier consideration of this relationship in a flock higli in liroodiness and 

 in a flock low in broodiness .seems advisable. The 1916 flock showed 86.31 per 

 cent broody and is unimproved, at least for broodiness. The 1923 flock siiowed 

 27.35 per cent broody and may be classified as an improved flock. 



4- Correlation Between the Presence of Broodiness and December Rate above 

 the Mean of Broodies and Non-broodies Combined — Pullet Year (Unimproved 

 Flock 1916). 



In the total of 253 birds the followiixg results were obtained: — 



