ANIMAL PRODUCTION. 669 



necessarily the stronger hen at all times of the year; and [results indicate] 

 that in selecting' for breeding stock to produce a large nun)l)er of chicl<ens and 

 capable pullets, the first essential is to select according to strength, with the 

 exix'ctatlou that hens so selected will usually be the most i)roductive." 



A biometrical study of egg- production in the domestic fowl. — III, Varia- 

 tion and correlation in the physical characters of the egg, 11. Pearl and F. 

 M. Surface ( L^. ,S'. Dept. Agr., Bur. Anim. Imlus. BuL 110, pi. 3 (1913), pp. 

 171-2^1, figs. 14). — This bulletin is a continuation of work previously noted 

 (E. S. R., 24, i>. 675), and repoi-ts the results of a detailed study made of the 

 normal variation and correlation of the egg of the domestic fowl in respect pri- 

 marily to si5!e and shape. While the studies were made with the egg of the 

 Barred riymouth Koek, it is thought probable that many of the results found 

 for the egg of this breed will also obtain for the eggs of other ])reeds. The 

 authors summarize their results as follows: 



" The egg is, to a signilicant degree, relatively more variable iu length than 

 in breadth. The egg is considerably more variable in shape, as measured by the 

 length-breadth index, than it is iu either of the linear dimensions length or 

 breadth. The weight of tlie egg and its volume (determined from length and 

 breadth on the assumjition that the egg is a perfect prolate siiheroid) are much 

 more variable than any of the otlier characters. There is a close agreement 

 between the egg of the domestic fowl and the human skull iu respect to the 

 degree of variation exhibited in the i'orresi>onding size and shai>e characters of 

 the two structures. A consideration of the processes concerned in the produc- 

 tion of the definitive size and shape of the egg and the skull suggests, though it 

 does not prove that this similarity with reference to variation constants is not 

 fortuitous, but rests upon a basis of general similarity iu regard to mechanical 

 factors operative in both cases during the development and fixation of the final 

 form. 



" With the exception of breadth, all other dimensional characters studied 

 show significantly skew (asymmetrical) variation. The difl'erent egg dimen- 

 sions give frequency distributions following different ones of Pearson's curve 

 types. Variation of the egg iu length and in shape (length-breadth index) fol- 

 lows a curve of type IV ; variation in breadth type II, and variation in volume 

 type I. 



"All of the egg dimensions studied are positively correlated with each other 

 iu varying degree. Intraracially egg length and egg breadth are correlated only 

 slightly. Furthermore, neither the weight nor the volume of the egg is more 

 than very slightly, if significantly at all, correlated with its shape. As would 

 be expected, both length and breadth are significantly correlated with the 

 weight and the volume of the egg. The bulk measures (volume and weight) are 

 more highly correlated with breadth than with length. There is a very high 

 correlation between volume and weight. This implies what was found in fact 

 to be the case, that the specific gravity of the egg exhibits only a very small 

 degree of variation. Thei*e is no significiiut net correlation between the shape 

 of an egg and its absolute size. The skull and the egg show a correspondence in 

 respect to the degree to which their chief dimensions are correlated. The cor- 

 relation results parallel those set forth above for simple variation. 



"A detailed study of the regression of the weight of the egg on length and 

 bieadth results in certain ecpiations from which the former quantity can be 

 estimated from a knowledge of the linear dimensions of the egg with a consid- 

 erable degree of accuracy. A comparison of the egg of the domestic fowl with 

 those of a number of species of wild birds in respect to relative variability 

 leads to the result that in general the eggs of wild birds are neither less nor 

 more variable than those of the domestic bird. It is concluded that probably 



