Of the many unusual morphological features that characterize mammals, perhaps the most intriguing and least understood is the enlarged brain. Among mammals, brain weight and body weight are allometrically related and follow the equation brain weight = k(body weight)a, where k is the allometric coefficient and a is the allometric exponent. The allometric exponent (regression coefficient) on a double logarithmic plot of brain weight on body weight across the Mammalia is estimated to be .74 (Eisenberg and Wilson, 1978; Martin, 1981; Armstrong, 1983); that is, larger mammals tend to have proportionately smaller brains. However, within groups of mammals, certain lineages show unusually enlarged brains; in such cases, brain mass far exceeds that predicted based on body mass alone. Examples of highly encephalized species are found in disparate mammalian lineages and at different taxonomic levels: pteropodid bats and humans exhibit greater encephalization than do other families in their respective orders (Jerison, 1973; Eisenberg and Wilson, 1978); certain porpoises and dolphins (Delphinidae), including Tursiops, Grampus, and Orcinus, stand out as the most encephalized cetaceans (Jerison, 1973, 1981; Osborne and Sundsten, 1981); opossums of the genus Caluromys seem to have larger brains in proportion to body size than do other didelphid marsupials (Eisenberg and Wilson, 1981); and relative brain size in certain species of white-footed mice (e.g., Peromyscus crinitus, P. truei, and P. pectoralis) exceeds that seen in other species of the genus (Mace and Eisenberg, 1982). Hypotheses designed to account for these unusually encephalized species of mammals are of two basic types. The first, which we term specialization, is founded on the assumption that increased specialization of a function (motor, sensory, or cognitive) requires increased brain mass to control that function (the principle of proper mass; Jerison, 1973). According to this view, relative brain size is a reliable index of both a species' trophic behavior and the complexity of the species' habitat (Findley and Wilson, 1982). This hypothesis directs one to search for ecoethological specializations in largebrained species to better understand the evolution of encephalization (for examples see Mann, 1963; Stephan and Pirlot, 1970; Pirlot and Pottier, 1977; Eisenberg and Wilson, 1978, 1981; Clutton-Brock and Harvey, 1980; Lemen, 1980; Mace et al., 1981; Roth and Thorington, 1982; Meier, 1983). The second hypothesis, articulated most recently by Gould (1977), suggests that the unusually enlarged brain may have originated through time shifts in an ancestral developmental sequence (heterochrony) resulting in a paedomorphic descendant possessing the proportionately large brain characteristic of the juvenile ancestor. According to the heterochrony hypothesis, then, one might search for evidence of ontogenetic perturbations in large-brained species to better understand the evolution of brain size. It is important to observe that these ideas are not directly competitive. The eco-ethological specialization hypothesis focuses on present use of the enlarged brain (with extrapolations as to its ori-