58 BULLETIN 1090, U. S. DEPARTMENT OF AGRICULTURE. 



the origin of tlie monsters. Inbreeding seems merely to have brougl 

 to light genetic traits in the original stock. 



Although most of the families came from the same line-bred stock 

 a striking differentiation with respect to traits connected with vigo 

 was found among them. These traits included size and frequency o 

 litters, percentage born alive and raised of those born alive, birtj 

 weight, and gain to 33 days. The differences between the familie 

 were greater than could be due to chance, and increased as the in 

 breeding progressed and the families became more homogeneou 

 through the elimination of early branches from the main lines o 

 descent. The families tended to keep the same rank with respect t( 

 each character. The correlation between the average grades in th( 

 early and later histories of the families was high in respect to siz( 

 of litter, birth weight, and gain. It was high enough to be significan 

 in the case of the percentage born alive. The correlations wer( 

 positive but of doubtful significance for the percentage rrised of th( 

 young born alive and for the frequency of litters. A detailed stud} 

 of the individual families, however, showed that the correlations woulc 

 have been higher and all would have been significant but for two or 

 three families in which there had been reversal of relative importance 

 among the subfamilies and in which therefore a change in rank in al] 

 or many respects was not surprising. Recent evidence indicates; 

 that even these families have now become fixed in their characteris-i 

 tics. The conclusion seems warranted that there was heredity of 

 all of the traits studied. 



There did not, however, appear to be heredity of general vigor.l 

 The average vigor of a family in one respect was found to be in the 

 main independent of its vigor in other respects. Thus the average 

 success of the families in raising their young was not correlated with" 

 weight or with size or frequency of litters. Neither was weight cor- 

 related with regularity in producing litters. There was not even a^ 

 significant correlation between the percentage born alive and the 

 percentage of those raised, although success or failure in each separa- 

 ately was undoubtedly characteristic of families. Similarly there was; 

 no correlation between the average size of litter and litter frequency. 

 The only apparent exception, outside of high correlations between 

 birth weight, gain, and year weight, was in a high correlation between! 

 weight and size of litter, for which there is undoubtedly some indi- 

 cation of a physiological interrelation. 



The study of the individual families brought out interesting ex- 

 amples of extreme vigor throughout the history of a family in certain 

 respects, associated with extreme weakness in others, as well as cases 

 in which all kinds of vigor or all kinds of weakness were combined. 



