Kluge and Kerfoot (1973) reported from several vetebrate data sets that interlocality character differentiation was positively correlated with intralocality variability for the same characters. Their method of analysis is criticized since it would give functional dependence in the null case of no geographic distribution. An alternative method of analysis is proposed based on coefficients of variation among and within localities. When Kluge and Kerfoot''s data sets are reexamined by this method, the relationship is not as high or as regular as reported by these authors. The 1st invertebrate data (from aphids, flies, snails and protozoa) are analyzed and yield results comparable to the finding on vertebrates.