EFFECTS OF INBREEDING AND CROSSBREEDING. 17 



shown above. Most of the variation within litters thus must be 

 due to such a cause as position in the uterus. 



Seasonal conditions have a marked effect on birth weight through 

 their influence on the condition of the dam. The age of the dam is 

 not an important factor. 



The gain in weight between birth and 1 33 days is affected by the 

 same factors which affect the prenatal rate of growth, namely, size 

 of litter, condition of dam, and heredity. 



The correlation between gain and size of litter in the same data as 

 those used above was —0.381 ±0.011. That between gain and inter- 

 val was +0.283 ±0.018. There is naturally considerable correla- 

 tion between birth weight and gain, the figure being +0.533 ±0.009. 

 Heredity plays a larger part than in the case of birth weight, and the 

 condition of the dam a smaller one. This was shown, as in the case 

 of the percentages born dead and reared, by the relatively greater 

 improvement in gain than in birth weight among the offspring of 

 inbred dams when mated with unrelated males instead of with 

 brothers. 



Hereditary variation, however, is not of much importance within 

 the control stock. The correlation between the mean gain of suc- 

 cessive litters was +0.221 ±0.026, and that of litters which were not 

 successive was +0.062 ±0.016. Common environmental factors 

 may play a part in the former case as well as heredity. 



Comparison of the gains made in different seasons and years demon- 

 strate the importance of differences in environmental conditions in 

 the present experiments. 



The effect of size of litter on the early growth is soon lost. There 

 is no correlation between size of litter and weight at a year among 

 the inbreds (r= +0.010 ±0.018, in males; r= -0.029 ±0.019 in 

 females) . The common influence of heredity, however, is seen in the 

 correlations of +0.395 ±0.031, +0.375 ±0.024, +0.374 ±0.035, 

 and +0.494 ±0.051, between weight at birth and weight at a year in 

 litters of 2 to 5 respectively among males. The correlations were 

 smaller among females ( + 0.339 ±0.032, +0.160 ±0.029, +0.187 

 ±0.041, +0.134 ±0.070), owing, doubtless, to the greater unrelia- 

 bility of their weights at a year. It was shown in an earlier paper 2 

 that the correlation of +0.375 ±0.024 for males in litters of 3 was in 

 part due to a correlation of +0.630 ±0.083 between average birth 

 weight and average year weight among 24 inbred families (including 

 Family 4) and in part to an average correlation of +0.308 ±0.026 

 between birth weight and year weight within a family. The value 



8 Wright, S. 1917. The average correlation -within subgroups of a population. Proc. Wash. Acad. Sci. 

 v. 7, p. 532-535. 



106851°— 22— Bull. 1090 3 



