Abstract
A coefficient of distance between taxa is proposed as an alternative measure of resemblance or similarity in numerical taxonomy. This distance can be computed as the mean sum of squares of the differences between standardized character state codes of the taxa concerned.