Allometry in primates, with emphasis on scaling and the evolution of the brain.

Allometry should be defined broadly as the study of size and its consequences, not narrowly as the application of power functions to the data of growth. Variation in size may be ontogenetic, static or phyletic. Errors of omission and treatment have plagued the study of allometry in primates. Standard texts often treat brain size as an independent measure, ignoring its allometric relation with body size - on this basis, gracile australopithecines have been accorded the mental status of gorillas. Intrinsic allometries of the brain/body are likewise neglected: many authors cite cerebral folding as evidence of man's mental superiority, but folding is a mechanical correlate of brain size itself. Confusion among types of scaling heads errors of treatment in both historical primacy [Dubois' ontogenetic inferences from interspecific curves] and current frequency. The predicted parameters of brain-body plots differ greatly for ontogenetic, intrapopulational, interspecific and phyletic allometries. I then discuss basic trends in bivariate allometry at the ordinal level for internal organ weights, skeletal dimensions, lifespan and fetal weight. In considering the causes of basic bivariate allometries, I examine the reason for differences among types of scaling in brain-body relationships. The interspecific exponent of 0.66 strongly suggests a relationship to body surfaces, but we have no satisfactory explanation for why this should be so. The tripartite ontogenetic plot is a consequence of patterns in neuronal differentiation. We do not know why intraspecific exponents fall between 0.2 and 0.4; several partial explanations have been offered. Multivariate techniques have transcended the pictorial representation of transformed coordinates and offer new, powerful approaches to total allometric patterns. Allometry is most often used as a 'criterion for subtraction'. In order to assess the nature and purpose of an adaptation, we must be able to identify and isolate the aspect of its form that depends both upon its size and the size of the body within which it resides. Cranial indices and limb lengths are misinterpreted when authors apply no correction for body size. The search for a criterion of subtraction has been most diligently pursued in studies of the brain. Clearly, brain size must be assessed by comparison with a 'standard' animal of the same body size. But how shall size be measured, especially in fossils; and how shall a standard animal be construed. I discuss and criticize three methods recently used: RADINSKY'S foramen magnum criterion; Jerison's minimum convex polygons and cephalization quotients; and the indices of progression in comparison with 'basal' insectivores' of BAUCHOT, Stephan and their colleagues.