In an ANOVA the term Sum of Squares refers to an interim quantity used in the calculation of an estimate of the population variance.

In a one-way ANOVA the Within Sum of Squares is a quantity used in the calculation of the Within Mean Square. More specifically the Within Sum of Squares divided by its Degrees of Freedom is equal to the Within Mean Square. Under the Null Hypothesis, the Within Mean Square is an unbiased estimate of the population variance.

Similarly, the Between Sum of Squares is a quantity used in the calculation of the Between Mean Square. More specifically the Between Mean Square is equal to the Between Sum of Squares divided by its Degrees of Freedom. Under the Null Hypothesis, the Between Mean Square is also an unbiased estimate of the population variance.