This paper by R. A. Fisher explores the correlation between relatives under the assumption of Mendelian inheritance. Fisher examines how genetic factors contribute to the variability observed in human traits, such as stature, and how these factors are distributed among different relatives. He introduces the concept of variance and shows how it can be partitioned into contributions from different genetic factors. Fisher also discusses the effects of dominance, epistasis, and assortative mating on the correlation between relatives. Fisher begins by considering the distribution of genetic factors in a population and how they contribute to the variance in traits. He shows that when two independent causes of variability act together, the resulting variance is the square root of the sum of their individual variances. He then considers the correlation between parents and children, showing that the variance in children can be explained by the variance in their parents, with the remaining variance attributed to other factors. Fisher also discusses the correlation between siblings, showing that the variance in siblings is partly due to shared ancestry. He introduces the concept of dominance, where one allele may mask the expression of another, and shows how this affects the correlation between relatives. He also considers the effects of epistasis, where the expression of one gene may depend on the presence of another gene. Fisher then extends his analysis to consider the correlation between different relatives, such as uncles and cousins, and shows how the correlation between these relatives is affected by the number of generations separating them. He also considers the effects of assortative mating, where individuals are more likely to mate with others who share similar traits, and shows how this affects the correlation between relatives. Finally, Fisher discusses the implications of multiple allelomorphism, where a single gene may have multiple alleles, and how this affects the correlation between relatives. He shows that even with multiple alleles, the correlation between relatives remains relatively simple, with the correlation between siblings still being exactly 1/4. The paper concludes with a discussion of the statistical implications of these findings for understanding human variability and the role of genetics in shaping it.This paper by R. A. Fisher explores the correlation between relatives under the assumption of Mendelian inheritance. Fisher examines how genetic factors contribute to the variability observed in human traits, such as stature, and how these factors are distributed among different relatives. He introduces the concept of variance and shows how it can be partitioned into contributions from different genetic factors. Fisher also discusses the effects of dominance, epistasis, and assortative mating on the correlation between relatives. Fisher begins by considering the distribution of genetic factors in a population and how they contribute to the variance in traits. He shows that when two independent causes of variability act together, the resulting variance is the square root of the sum of their individual variances. He then considers the correlation between parents and children, showing that the variance in children can be explained by the variance in their parents, with the remaining variance attributed to other factors. Fisher also discusses the correlation between siblings, showing that the variance in siblings is partly due to shared ancestry. He introduces the concept of dominance, where one allele may mask the expression of another, and shows how this affects the correlation between relatives. He also considers the effects of epistasis, where the expression of one gene may depend on the presence of another gene. Fisher then extends his analysis to consider the correlation between different relatives, such as uncles and cousins, and shows how the correlation between these relatives is affected by the number of generations separating them. He also considers the effects of assortative mating, where individuals are more likely to mate with others who share similar traits, and shows how this affects the correlation between relatives. Finally, Fisher discusses the implications of multiple allelomorphism, where a single gene may have multiple alleles, and how this affects the correlation between relatives. He shows that even with multiple alleles, the correlation between relatives remains relatively simple, with the correlation between siblings still being exactly 1/4. The paper concludes with a discussion of the statistical implications of these findings for understanding human variability and the role of genetics in shaping it.