50 



BULLETIN 1090, U. S. DEPARTMENT OF AGRICULTURE. 



rank C; the next five, rank D; and the poorest three, rank E. Family 

 15, which produced too few young in the second period to give 

 significant results, is assigned ranks in the first period but is not 

 counted in assigning ranks to the other families. 



Table 5. — Ranis of the families during years 1906 to 1910, and 1911 to 1915. 



[The best three are given rank A; the next five, rank B ; the middle six, rank C; the next five, rank D; 

 and the poorest three, rank E. Family 15 is given a rank in the first period onlv and is not counted 

 in assigning ranks to the others. The ranks for 1906-1910 and 1911-1915 are given in succession. For 

 example, EC means among the poorest three in 1906-1910 and middling in 1911-1915. Great changes 

 in rank are indicated by stars.] 



Family. 



Per cent 

 born 

 ! alive. 



Per cent 



raised of 



those 



born 



alive. 



Per cent 

 raised. 



Birth 

 weight 

 (those 

 raised). 



Gain to 

 33 days. 



Weight 

 at 33 

 days. 



Size of 

 fitter. 



Litters 

 per year. 



Young 

 per year. 



1 



cc 



EE 



ED 



ED 



DE 



EE 



EE 



DE 



EE 



2 



EC* 



CA* 



DB* 



CE* 



EE 



EE 



DD 



EA*** 



EB** 



3 



EE 



CE* 



EE 



BC 



BA 



BB 



BC 



DB* 



CB 



7 



DD 



DB* 



EC* 



DC 



CC 



CB 



BC 



CC 



BB 



9 



DD 



CC 



DD 



AB 



AA 



AA 



BA 



CD 



BC 



11 



DC 



EE 



DE 



BA 



BA 



BA 



AA 



DD 



BC 



13 



CB 



CB 



CB 



AA 



AB 



AA 



AB 



CC 



BA 



14 



EE 



CD 



DD 



CC 



CB 



CB 



CC 



CE* 



CE* 



15 



E 



B 



E 



C 



E 



D 



E 



D 



E 



17 



AB 



AB 



AA 



ED 



ED 



ED 



DD 



DC 



DC 



18 



BB 



DD 



BC 



BB 



CC 



CC 



CC 



BB 



CC 



19 



DD 



BD* 



CD 



CB 



DD 



DC 



BB 



BB 



BA 



20 



DB* 



DA** 



DA** 



CC 



CD 



CD 



- EC* 



CD 



DD 



21 



BD* 



DB* 



CC 



AD** 



BC 



AC* 



EE 



BB 



DD 



23 



AA 



ED 



CC 



BC 



BD* 



BE** 



BD* 



AA 



AB 



24 



CE* 



BD* 



BE** 



DE 



DC 



DC 



CE* 



AB 



AD** 



31 



AD**' 



BA 



AB 



CC 



DC 



CC 



CB 



ED 



DD 



32 



BC 



CC 



BD* 



DD 



ED 



DD 



CD 



BC 



CC 



34 



CB 



AC* 



BC 



BB 



BE** 



BD* 



DD 



CD 



CD 



35 



BC 



DB* 



CB 



DB* 



DB* 



DB* 



DA** 



DA** 



DA** 



36 



CC 



BC 



BC 



DD 



CB 



CC 



AC* 



BC 



AC* 



38 



BA 



•AC* 



AB 



CA* 



AB 



BB 



CB 



EE 



EE 



39 



CA* 



BC 



CA* 



EE 



CC 



DD 



DB* 



AC* 



CB 



Inspection of Table 5 brings out clearly the tendency of the fam- 

 ilies to hold their rank with respect to each character. Changes of 

 rank of more than one grade are indicated by stars. A shift of two 

 grades is shown by one star, of three grades by two stars, and one 

 casein which there was a change from E to A by three stars. There 

 was the greatest shifting of places in the percentage raised among 

 those born alive, but even here 12 of the 22 held their rank or changed 

 by only one grade. The correlation between successive periods 

 was rather low in this case, being only +0.32. In the cases of the 

 other characters, there were from 16 to 19 of the 22 families which 

 held their place in the sense given above or changed only one grade. 

 The case of the frequency of litters is especially interesting, since the 

 coefficient of correlation between successive periods ( + 0.25) did not 

 appear to be significant. It is therefore surprising to find that eight 

 of the families were of the same grade in both periods and that nine 

 changed by only one grade, leaving only fixe which made conspicuous 

 changes in rank. The low correlation is evidently due to the remark- 

 able change in character of two families, namely, Nos. 2 and 35. 



