Imagine the individuals have been measured for two characters, X and Y. The covariance between the two is defined as
covXY = (1/ (n - 1)) ∑ (Xi - Xm) (Yi - Ym)
Covariance measures whether, if an individual has a large value of X it also has a large value for Y. If the Xi and Yi of an individual are both large, the product XiYi will also be large; if Yi is not large when Xi is, the product will be smaller. Generally, if X and Y covary, the product (and so the covariance) is large; and if they do not, the sum of the products will come to zero. See also variance.